Update index.html
Browse files- index.html +1883 -18
index.html
CHANGED
|
@@ -1,19 +1,1884 @@
|
|
| 1 |
-
<!
|
| 2 |
-
<html>
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
|
| 11 |
-
|
| 12 |
-
|
| 13 |
-
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
|
| 17 |
-
|
| 18 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
</html>
|
|
|
|
| 1 |
+
<!DOCTYPE html>
|
| 2 |
+
<html lang="en">
|
| 3 |
+
<head>
|
| 4 |
+
<meta charset="UTF-8">
|
| 5 |
+
<meta name="viewport" content="width=device-width, initial-scale=1.0">
|
| 6 |
+
<title>DSMOTE — Interactive Visualization</title>
|
| 7 |
+
<link href="https://fonts.googleapis.com/css2?family=Rajdhani:wght@400;500;600;700&family=JetBrains+Mono:wght@300;400;500&family=Syne:wght@400;700;800&display=swap" rel="stylesheet">
|
| 8 |
+
<style>
|
| 9 |
+
:root {
|
| 10 |
+
--bg: #060b14;
|
| 11 |
+
--bg2: #0d1627;
|
| 12 |
+
--bg3: #132038;
|
| 13 |
+
--cyan: #00e5ff;
|
| 14 |
+
--cyan2: #00bcd4;
|
| 15 |
+
--orange: #ff6b35;
|
| 16 |
+
--green: #00e676;
|
| 17 |
+
--red: #ff1744;
|
| 18 |
+
--purple: #7c4dff;
|
| 19 |
+
--yellow: #ffd740;
|
| 20 |
+
--text: #e0f7fa;
|
| 21 |
+
--muted: #546e7a;
|
| 22 |
+
--border: rgba(0,229,255,0.15);
|
| 23 |
+
}
|
| 24 |
+
|
| 25 |
+
* { box-sizing: border-box; margin: 0; padding: 0; }
|
| 26 |
+
|
| 27 |
+
body {
|
| 28 |
+
background: var(--bg);
|
| 29 |
+
color: var(--text);
|
| 30 |
+
font-family: 'Rajdhani', sans-serif;
|
| 31 |
+
min-height: 100vh;
|
| 32 |
+
overflow-x: hidden;
|
| 33 |
+
}
|
| 34 |
+
|
| 35 |
+
/* Grid noise overlay */
|
| 36 |
+
body::before {
|
| 37 |
+
content: '';
|
| 38 |
+
position: fixed;
|
| 39 |
+
inset: 0;
|
| 40 |
+
background-image:
|
| 41 |
+
linear-gradient(rgba(0,229,255,0.03) 1px, transparent 1px),
|
| 42 |
+
linear-gradient(90deg, rgba(0,229,255,0.03) 1px, transparent 1px);
|
| 43 |
+
background-size: 40px 40px;
|
| 44 |
+
pointer-events: none;
|
| 45 |
+
z-index: 0;
|
| 46 |
+
}
|
| 47 |
+
|
| 48 |
+
.container { max-width: 1100px; margin: 0 auto; padding: 0 24px; position: relative; z-index: 1; }
|
| 49 |
+
|
| 50 |
+
/* HEADER */
|
| 51 |
+
header {
|
| 52 |
+
padding: 36px 0 24px;
|
| 53 |
+
border-bottom: 1px solid var(--border);
|
| 54 |
+
margin-bottom: 32px;
|
| 55 |
+
}
|
| 56 |
+
.header-tag {
|
| 57 |
+
font-family: 'JetBrains Mono', monospace;
|
| 58 |
+
font-size: 11px;
|
| 59 |
+
color: var(--cyan);
|
| 60 |
+
letter-spacing: 3px;
|
| 61 |
+
text-transform: uppercase;
|
| 62 |
+
margin-bottom: 10px;
|
| 63 |
+
opacity: 0.7;
|
| 64 |
+
}
|
| 65 |
+
header h1 {
|
| 66 |
+
font-family: 'Syne', sans-serif;
|
| 67 |
+
font-size: 2.4rem;
|
| 68 |
+
font-weight: 800;
|
| 69 |
+
letter-spacing: -1px;
|
| 70 |
+
background: linear-gradient(135deg, var(--cyan) 0%, #ffffff 60%);
|
| 71 |
+
-webkit-background-clip: text;
|
| 72 |
+
-webkit-text-fill-color: transparent;
|
| 73 |
+
background-clip: text;
|
| 74 |
+
}
|
| 75 |
+
header p {
|
| 76 |
+
font-family: 'JetBrains Mono', monospace;
|
| 77 |
+
font-size: 12px;
|
| 78 |
+
color: var(--muted);
|
| 79 |
+
margin-top: 8px;
|
| 80 |
+
letter-spacing: 1px;
|
| 81 |
+
}
|
| 82 |
+
|
| 83 |
+
/* TABS */
|
| 84 |
+
.tabs {
|
| 85 |
+
display: flex;
|
| 86 |
+
gap: 4px;
|
| 87 |
+
margin-bottom: 28px;
|
| 88 |
+
flex-wrap: wrap;
|
| 89 |
+
}
|
| 90 |
+
.tab {
|
| 91 |
+
font-family: 'JetBrains Mono', monospace;
|
| 92 |
+
font-size: 11px;
|
| 93 |
+
padding: 8px 16px;
|
| 94 |
+
border: 1px solid var(--border);
|
| 95 |
+
background: transparent;
|
| 96 |
+
color: var(--muted);
|
| 97 |
+
cursor: pointer;
|
| 98 |
+
letter-spacing: 1px;
|
| 99 |
+
text-transform: uppercase;
|
| 100 |
+
transition: all 0.2s;
|
| 101 |
+
position: relative;
|
| 102 |
+
}
|
| 103 |
+
.tab:hover { color: var(--cyan); border-color: var(--cyan); }
|
| 104 |
+
.tab.active {
|
| 105 |
+
background: var(--cyan);
|
| 106 |
+
color: var(--bg);
|
| 107 |
+
border-color: var(--cyan);
|
| 108 |
+
font-weight: 500;
|
| 109 |
+
}
|
| 110 |
+
.tab-num {
|
| 111 |
+
display: inline-block;
|
| 112 |
+
margin-right: 6px;
|
| 113 |
+
opacity: 0.5;
|
| 114 |
+
}
|
| 115 |
+
|
| 116 |
+
/* PANELS */
|
| 117 |
+
.panel { display: none; animation: fadeIn 0.3s ease; }
|
| 118 |
+
.panel.active { display: block; }
|
| 119 |
+
@keyframes fadeIn { from { opacity: 0; transform: translateY(8px); } to { opacity: 1; transform: none; } }
|
| 120 |
+
|
| 121 |
+
/* CARD */
|
| 122 |
+
.card {
|
| 123 |
+
background: var(--bg2);
|
| 124 |
+
border: 1px solid var(--border);
|
| 125 |
+
border-radius: 2px;
|
| 126 |
+
padding: 28px;
|
| 127 |
+
margin-bottom: 20px;
|
| 128 |
+
}
|
| 129 |
+
.card-title {
|
| 130 |
+
font-family: 'Syne', sans-serif;
|
| 131 |
+
font-size: 1.3rem;
|
| 132 |
+
font-weight: 700;
|
| 133 |
+
color: var(--cyan);
|
| 134 |
+
margin-bottom: 6px;
|
| 135 |
+
}
|
| 136 |
+
.card-sub {
|
| 137 |
+
font-family: 'JetBrains Mono', monospace;
|
| 138 |
+
font-size: 11px;
|
| 139 |
+
color: var(--muted);
|
| 140 |
+
margin-bottom: 20px;
|
| 141 |
+
letter-spacing: 1px;
|
| 142 |
+
}
|
| 143 |
+
|
| 144 |
+
/* CANVAS */
|
| 145 |
+
canvas {
|
| 146 |
+
display: block;
|
| 147 |
+
border: 1px solid var(--border);
|
| 148 |
+
border-radius: 2px;
|
| 149 |
+
background: var(--bg);
|
| 150 |
+
}
|
| 151 |
+
|
| 152 |
+
.plot-row {
|
| 153 |
+
display: grid;
|
| 154 |
+
grid-template-columns: 1fr 1fr;
|
| 155 |
+
gap: 20px;
|
| 156 |
+
align-items: start;
|
| 157 |
+
}
|
| 158 |
+
|
| 159 |
+
.plot-label {
|
| 160 |
+
font-family: 'JetBrains Mono', monospace;
|
| 161 |
+
font-size: 11px;
|
| 162 |
+
text-align: center;
|
| 163 |
+
margin-top: 8px;
|
| 164 |
+
letter-spacing: 1px;
|
| 165 |
+
text-transform: uppercase;
|
| 166 |
+
}
|
| 167 |
+
|
| 168 |
+
/* BTN */
|
| 169 |
+
.btn {
|
| 170 |
+
font-family: 'JetBrains Mono', monospace;
|
| 171 |
+
font-size: 11px;
|
| 172 |
+
padding: 10px 20px;
|
| 173 |
+
border: 1px solid var(--cyan);
|
| 174 |
+
background: transparent;
|
| 175 |
+
color: var(--cyan);
|
| 176 |
+
cursor: pointer;
|
| 177 |
+
letter-spacing: 2px;
|
| 178 |
+
text-transform: uppercase;
|
| 179 |
+
transition: all 0.2s;
|
| 180 |
+
margin-top: 16px;
|
| 181 |
+
margin-right: 8px;
|
| 182 |
+
}
|
| 183 |
+
.btn:hover { background: var(--cyan); color: var(--bg); }
|
| 184 |
+
.btn-orange { border-color: var(--orange); color: var(--orange); }
|
| 185 |
+
.btn-orange:hover { background: var(--orange); color: var(--bg); }
|
| 186 |
+
|
| 187 |
+
/* LEGEND */
|
| 188 |
+
.legend {
|
| 189 |
+
display: flex;
|
| 190 |
+
gap: 20px;
|
| 191 |
+
flex-wrap: wrap;
|
| 192 |
+
margin-top: 14px;
|
| 193 |
+
}
|
| 194 |
+
.legend-item {
|
| 195 |
+
display: flex;
|
| 196 |
+
align-items: center;
|
| 197 |
+
gap: 8px;
|
| 198 |
+
font-family: 'JetBrains Mono', monospace;
|
| 199 |
+
font-size: 11px;
|
| 200 |
+
color: var(--muted);
|
| 201 |
+
letter-spacing: 0.5px;
|
| 202 |
+
}
|
| 203 |
+
.legend-dot {
|
| 204 |
+
width: 10px; height: 10px; border-radius: 50%;
|
| 205 |
+
flex-shrink: 0;
|
| 206 |
+
}
|
| 207 |
+
.legend-line {
|
| 208 |
+
width: 20px; height: 2px;
|
| 209 |
+
flex-shrink: 0;
|
| 210 |
+
}
|
| 211 |
+
|
| 212 |
+
/* PIPELINE */
|
| 213 |
+
.pipeline {
|
| 214 |
+
display: flex;
|
| 215 |
+
align-items: center;
|
| 216 |
+
gap: 0;
|
| 217 |
+
margin: 24px 0;
|
| 218 |
+
overflow-x: auto;
|
| 219 |
+
padding-bottom: 10px;
|
| 220 |
+
}
|
| 221 |
+
.pipe-step {
|
| 222 |
+
display: flex;
|
| 223 |
+
flex-direction: column;
|
| 224 |
+
align-items: center;
|
| 225 |
+
gap: 10px;
|
| 226 |
+
flex-shrink: 0;
|
| 227 |
+
opacity: 0;
|
| 228 |
+
transform: translateY(20px);
|
| 229 |
+
transition: opacity 0.5s ease, transform 0.5s ease;
|
| 230 |
+
}
|
| 231 |
+
.pipe-step.visible { opacity: 1; transform: none; }
|
| 232 |
+
.pipe-icon {
|
| 233 |
+
width: 70px; height: 70px;
|
| 234 |
+
border-radius: 2px;
|
| 235 |
+
display: flex;
|
| 236 |
+
align-items: center;
|
| 237 |
+
justify-content: center;
|
| 238 |
+
font-size: 26px;
|
| 239 |
+
border: 2px solid;
|
| 240 |
+
position: relative;
|
| 241 |
+
cursor: pointer;
|
| 242 |
+
}
|
| 243 |
+
.pipe-icon::after {
|
| 244 |
+
content: '';
|
| 245 |
+
position: absolute;
|
| 246 |
+
inset: -4px;
|
| 247 |
+
border-radius: 4px;
|
| 248 |
+
opacity: 0;
|
| 249 |
+
transition: opacity 0.3s;
|
| 250 |
+
}
|
| 251 |
+
.pipe-icon:hover::after { opacity: 1; }
|
| 252 |
+
.pipe-icon.active-step { box-shadow: 0 0 20px currentColor; }
|
| 253 |
+
|
| 254 |
+
.pipe-label {
|
| 255 |
+
font-family: 'JetBrains Mono', monospace;
|
| 256 |
+
font-size: 9px;
|
| 257 |
+
letter-spacing: 1px;
|
| 258 |
+
text-transform: uppercase;
|
| 259 |
+
text-align: center;
|
| 260 |
+
max-width: 80px;
|
| 261 |
+
}
|
| 262 |
+
.pipe-arrow {
|
| 263 |
+
font-size: 20px;
|
| 264 |
+
color: var(--muted);
|
| 265 |
+
padding: 0 6px;
|
| 266 |
+
flex-shrink: 0;
|
| 267 |
+
margin-top: -20px;
|
| 268 |
+
}
|
| 269 |
+
|
| 270 |
+
.step-detail {
|
| 271 |
+
background: var(--bg3);
|
| 272 |
+
border-left: 3px solid var(--cyan);
|
| 273 |
+
padding: 16px 20px;
|
| 274 |
+
margin-top: 8px;
|
| 275 |
+
border-radius: 0 2px 2px 0;
|
| 276 |
+
display: none;
|
| 277 |
+
}
|
| 278 |
+
.step-detail.visible { display: block; animation: fadeIn 0.3s; }
|
| 279 |
+
.step-detail h4 {
|
| 280 |
+
font-family: 'Syne', sans-serif;
|
| 281 |
+
font-size: 1.1rem;
|
| 282 |
+
margin-bottom: 6px;
|
| 283 |
+
}
|
| 284 |
+
.step-detail p {
|
| 285 |
+
font-family: 'JetBrains Mono', monospace;
|
| 286 |
+
font-size: 11px;
|
| 287 |
+
color: var(--muted);
|
| 288 |
+
line-height: 1.7;
|
| 289 |
+
}
|
| 290 |
+
.formula {
|
| 291 |
+
font-family: 'JetBrains Mono', monospace;
|
| 292 |
+
font-size: 12px;
|
| 293 |
+
background: var(--bg);
|
| 294 |
+
padding: 8px 14px;
|
| 295 |
+
margin-top: 10px;
|
| 296 |
+
border: 1px solid var(--border);
|
| 297 |
+
color: var(--cyan);
|
| 298 |
+
display: inline-block;
|
| 299 |
+
}
|
| 300 |
+
|
| 301 |
+
/* BEFORE/AFTER */
|
| 302 |
+
.ba-row {
|
| 303 |
+
display: grid;
|
| 304 |
+
grid-template-columns: 1fr 1fr;
|
| 305 |
+
gap: 24px;
|
| 306 |
+
}
|
| 307 |
+
.ba-card {
|
| 308 |
+
background: var(--bg3);
|
| 309 |
+
border: 1px solid var(--border);
|
| 310 |
+
padding: 20px;
|
| 311 |
+
border-radius: 2px;
|
| 312 |
+
}
|
| 313 |
+
.ba-card h4 {
|
| 314 |
+
font-family: 'Syne', sans-serif;
|
| 315 |
+
font-size: 1rem;
|
| 316 |
+
margin-bottom: 16px;
|
| 317 |
+
text-align: center;
|
| 318 |
+
}
|
| 319 |
+
.bar-row {
|
| 320 |
+
display: flex;
|
| 321 |
+
align-items: center;
|
| 322 |
+
gap: 10px;
|
| 323 |
+
margin-bottom: 9px;
|
| 324 |
+
}
|
| 325 |
+
.bar-name {
|
| 326 |
+
font-family: 'JetBrains Mono', monospace;
|
| 327 |
+
font-size: 9px;
|
| 328 |
+
width: 70px;
|
| 329 |
+
text-align: right;
|
| 330 |
+
color: var(--muted);
|
| 331 |
+
flex-shrink: 0;
|
| 332 |
+
letter-spacing: 0.5px;
|
| 333 |
+
overflow: hidden;
|
| 334 |
+
text-overflow: ellipsis;
|
| 335 |
+
white-space: nowrap;
|
| 336 |
+
}
|
| 337 |
+
.bar-track {
|
| 338 |
+
flex: 1;
|
| 339 |
+
height: 14px;
|
| 340 |
+
background: var(--bg);
|
| 341 |
+
border-radius: 1px;
|
| 342 |
+
overflow: hidden;
|
| 343 |
+
border: 1px solid rgba(255,255,255,0.05);
|
| 344 |
+
}
|
| 345 |
+
.bar-fill {
|
| 346 |
+
height: 100%;
|
| 347 |
+
border-radius: 1px;
|
| 348 |
+
transition: width 1s cubic-bezier(0.4,0,0.2,1);
|
| 349 |
+
position: relative;
|
| 350 |
+
}
|
| 351 |
+
.bar-val {
|
| 352 |
+
font-family: 'JetBrains Mono', monospace;
|
| 353 |
+
font-size: 9px;
|
| 354 |
+
width: 50px;
|
| 355 |
+
color: var(--muted);
|
| 356 |
+
flex-shrink: 0;
|
| 357 |
+
}
|
| 358 |
+
|
| 359 |
+
/* INSIGHT BOX */
|
| 360 |
+
.insight {
|
| 361 |
+
background: rgba(0,229,255,0.05);
|
| 362 |
+
border: 1px solid rgba(0,229,255,0.2);
|
| 363 |
+
padding: 14px 18px;
|
| 364 |
+
margin-top: 16px;
|
| 365 |
+
border-radius: 1px;
|
| 366 |
+
}
|
| 367 |
+
.insight-label {
|
| 368 |
+
font-family: 'JetBrains Mono', monospace;
|
| 369 |
+
font-size: 9px;
|
| 370 |
+
color: var(--cyan);
|
| 371 |
+
letter-spacing: 2px;
|
| 372 |
+
text-transform: uppercase;
|
| 373 |
+
margin-bottom: 6px;
|
| 374 |
+
}
|
| 375 |
+
.insight p {
|
| 376 |
+
font-family: 'Rajdhani', sans-serif;
|
| 377 |
+
font-size: 1rem;
|
| 378 |
+
font-weight: 600;
|
| 379 |
+
color: var(--text);
|
| 380 |
+
line-height: 1.5;
|
| 381 |
+
}
|
| 382 |
+
|
| 383 |
+
@media (max-width: 640px) {
|
| 384 |
+
.plot-row, .ba-row { grid-template-columns: 1fr; }
|
| 385 |
+
header h1 { font-size: 1.6rem; }
|
| 386 |
+
.pipeline { gap: 0; }
|
| 387 |
+
}
|
| 388 |
+
</style>
|
| 389 |
+
</head>
|
| 390 |
+
<body>
|
| 391 |
+
<div class="container">
|
| 392 |
+
<header>
|
| 393 |
+
<div class="header-tag">// Conference Visualization — DSMOTE</div>
|
| 394 |
+
<h1>Dynamic SMOTE: Interactive Visual Explainer</h1>
|
| 395 |
+
<p>A Hybrid Oversampling Framework for NIDS Class Imbalance</p>
|
| 396 |
+
</header>
|
| 397 |
+
|
| 398 |
+
<div class="tabs">
|
| 399 |
+
<button class="tab active" onclick="switchTab(0)"><span class="tab-num">01</span>SMOTE Weakness</button>
|
| 400 |
+
<button class="tab" onclick="switchTab(1)"><span class="tab-num">02</span>DSMOTE Pipeline</button>
|
| 401 |
+
<button class="tab" onclick="switchTab(2)"><span class="tab-num">03</span>Clustering</button>
|
| 402 |
+
<button class="tab" onclick="switchTab(3)"><span class="tab-num">04</span>Density Constraint</button>
|
| 403 |
+
<button class="tab" onclick="switchTab(4)"><span class="tab-num">05</span>Before vs After</button>
|
| 404 |
+
<button class="tab" onclick="switchTab(5)"><span class="tab-num">06</span>UNSW Results</button>
|
| 405 |
+
<button class="tab" onclick="switchTab(6)"><span class="tab-num">07</span>KDD Results</button>
|
| 406 |
+
</div>
|
| 407 |
+
|
| 408 |
+
<!-- ======================== PANEL 1: SMOTE WEAKNESS ======================== -->
|
| 409 |
+
<div class="panel active" id="panel-0">
|
| 410 |
+
<div class="card">
|
| 411 |
+
<div class="card-title">SMOTE is Blind</div>
|
| 412 |
+
<div class="card-sub">// Standard SMOTE interpolates between any two minority samples — crossing cluster boundaries</div>
|
| 413 |
+
<div class="plot-row">
|
| 414 |
+
<div>
|
| 415 |
+
<canvas id="smote-canvas" width="440" height="360"></canvas>
|
| 416 |
+
<div class="plot-label" style="color:var(--orange)">⚠ Standard SMOTE — Noise Generation</div>
|
| 417 |
+
</div>
|
| 418 |
+
<div>
|
| 419 |
+
<canvas id="dsmote-canvas" width="440" height="360"></canvas>
|
| 420 |
+
<div class="plot-label" style="color:var(--green)">✓ DSMOTE — Cluster-Aware Sampling</div>
|
| 421 |
+
</div>
|
| 422 |
+
</div>
|
| 423 |
+
<div style="margin-top:12px;">
|
| 424 |
+
<button class="btn" onclick="animateSMOTE()">▶ Animate SMOTE</button>
|
| 425 |
+
<button class="btn" onclick="animateDSMOTE()">▶ Animate DSMOTE</button>
|
| 426 |
+
<button class="btn btn-orange" onclick="resetSmote()">↺ Reset</button>
|
| 427 |
+
</div>
|
| 428 |
+
<div class="legend">
|
| 429 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--cyan)"></div>Cluster A — Minority</div>
|
| 430 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--purple)"></div>Cluster B — Minority</div>
|
| 431 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--orange);opacity:0.6"></div>SMOTE — Noise points (wrong region)</div>
|
| 432 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--green)"></div>DSMOTE — Safe synthetic points</div>
|
| 433 |
+
</div>
|
| 434 |
+
<div class="insight">
|
| 435 |
+
<div class="insight-label">// Key Message</div>
|
| 436 |
+
<p>SMOTE selects two minority samples at random — regardless of which cluster they belong to — and interpolates between them. This creates points in empty space between clusters, introducing noise and confusion for the classifier.</p>
|
| 437 |
+
</div>
|
| 438 |
+
</div>
|
| 439 |
+
</div>
|
| 440 |
+
|
| 441 |
+
<!-- ======================== PANEL 2: PIPELINE ======================== -->
|
| 442 |
+
<div class="panel" id="panel-1">
|
| 443 |
+
<div class="card">
|
| 444 |
+
<div class="card-title">DSMOTE Algorithm Pipeline</div>
|
| 445 |
+
<div class="card-sub">// Click each step to explore the details</div>
|
| 446 |
+
<div class="pipeline" id="pipeline">
|
| 447 |
+
<div class="pipe-step" id="pstep-0">
|
| 448 |
+
<div class="pipe-icon" onclick="showStep(0)"
|
| 449 |
+
style="color:#ff6b35;border-color:#ff6b35;background:rgba(255,107,53,0.08)">✂️</div>
|
| 450 |
+
<div class="pipe-label" style="color:#ff6b35">Majority Reduction</div>
|
| 451 |
+
</div>
|
| 452 |
+
<div class="pipe-arrow">→</div>
|
| 453 |
+
<div class="pipe-step" id="pstep-1">
|
| 454 |
+
<div class="pipe-icon" onclick="showStep(1)"
|
| 455 |
+
style="color:#7c4dff;border-color:#7c4dff;background:rgba(124,77,255,0.08)">📉</div>
|
| 456 |
+
<div class="pipe-label" style="color:#7c4dff">PCA Reduction</div>
|
| 457 |
+
</div>
|
| 458 |
+
<div class="pipe-arrow">→</div>
|
| 459 |
+
<div class="pipe-step" id="pstep-2">
|
| 460 |
+
<div class="pipe-icon" onclick="showStep(2)"
|
| 461 |
+
style="color:#00e5ff;border-color:#00e5ff;background:rgba(0,229,255,0.08)">🔵</div>
|
| 462 |
+
<div class="pipe-label" style="color:#00e5ff">KMeans Clustering</div>
|
| 463 |
+
</div>
|
| 464 |
+
<div class="pipe-arrow">→</div>
|
| 465 |
+
<div class="pipe-step" id="pstep-3">
|
| 466 |
+
<div class="pipe-icon" onclick="showStep(3)"
|
| 467 |
+
style="color:#ffd740;border-color:#ffd740;background:rgba(255,215,64,0.08)">⚡</div>
|
| 468 |
+
<div class="pipe-label" style="color:#ffd740">Smart Sampling</div>
|
| 469 |
+
</div>
|
| 470 |
+
<div class="pipe-arrow">→</div>
|
| 471 |
+
<div class="pipe-step" id="pstep-4">
|
| 472 |
+
<div class="pipe-icon" onclick="showStep(4)"
|
| 473 |
+
style="color:#00e676;border-color:#00e676;background:rgba(0,230,118,0.08)">🎯</div>
|
| 474 |
+
<div class="pipe-label" style="color:#00e676">Density Filter</div>
|
| 475 |
+
</div>
|
| 476 |
+
<div class="pipe-arrow">→</div>
|
| 477 |
+
<div class="pipe-step" id="pstep-5">
|
| 478 |
+
<div class="pipe-icon" onclick="showStep(5)"
|
| 479 |
+
style="color:#ff4081;border-color:#ff4081;background:rgba(255,64,129,0.08)">⚖️</div>
|
| 480 |
+
<div class="pipe-label" style="color:#ff4081">Class Weights</div>
|
| 481 |
+
</div>
|
| 482 |
+
</div>
|
| 483 |
+
|
| 484 |
+
<div id="step-detail" class="step-detail visible">
|
| 485 |
+
<h4 style="color:var(--cyan)">👆 Click a Step Above</h4>
|
| 486 |
+
<p>Each step in DSMOTE solves a specific problem. Click any step icon to learn what it does and why it matters.</p>
|
| 487 |
+
</div>
|
| 488 |
+
</div>
|
| 489 |
+
|
| 490 |
+
<!-- Mini pipeline canvas -->
|
| 491 |
+
<div class="card">
|
| 492 |
+
<div class="card-title">Pipeline Data Flow</div>
|
| 493 |
+
<div class="card-sub">// How raw data transforms through each stage</div>
|
| 494 |
+
<canvas id="pipeline-canvas" width="1040" height="200" style="width:100%"></canvas>
|
| 495 |
+
</div>
|
| 496 |
+
</div>
|
| 497 |
+
|
| 498 |
+
<!-- ======================== PANEL 3: CLUSTERING ======================== -->
|
| 499 |
+
<div class="panel" id="panel-2">
|
| 500 |
+
<div class="card">
|
| 501 |
+
<div class="card-title">Step 3 — KMeans Clustering of Minority Class</div>
|
| 502 |
+
<div class="card-sub">// DSMOTE understands the internal structure of minority classes</div>
|
| 503 |
+
<canvas id="cluster-canvas" width="700" height="420" style="width:100%"></canvas>
|
| 504 |
+
<div style="margin-top:12px">
|
| 505 |
+
<button class="btn" onclick="animateClusters()">▶ Show Clustering</button>
|
| 506 |
+
<button class="btn btn-orange" onclick="resetClusters()">↺ Reset</button>
|
| 507 |
+
</div>
|
| 508 |
+
<div class="legend" style="margin-top:14px">
|
| 509 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--cyan)"></div>Cluster 1</div>
|
| 510 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--orange)"></div>Cluster 2</div>
|
| 511 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--purple)"></div>Cluster 3</div>
|
| 512 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--yellow)"></div>Cluster 4</div>
|
| 513 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--muted)"></div>Uncolored (before clustering)</div>
|
| 514 |
+
</div>
|
| 515 |
+
<div class="insight">
|
| 516 |
+
<div class="insight-label">// Key Message</div>
|
| 517 |
+
<p>Instead of treating all minority samples as one blob, DSMOTE uses KMeans to discover sub-groups. Synthetic samples are then generated <em>within each cluster</em> — not across them.</p>
|
| 518 |
+
</div>
|
| 519 |
+
</div>
|
| 520 |
+
</div>
|
| 521 |
+
|
| 522 |
+
<!-- ======================== PANEL 4: DENSITY CONSTRAINT ======================== -->
|
| 523 |
+
<div class="panel" id="panel-3">
|
| 524 |
+
<div class="card">
|
| 525 |
+
<div class="card-title">Step 5 — Density Constraint Filtering</div>
|
| 526 |
+
<div class="card-sub">// Only synthetic points within the mean intra-cluster distance are accepted</div>
|
| 527 |
+
<canvas id="density-canvas" width="700" height="420" style="width:100%"></canvas>
|
| 528 |
+
<div style="margin-top:12px">
|
| 529 |
+
<button class="btn" onclick="animateDensity()">▶ Generate Samples</button>
|
| 530 |
+
<button class="btn btn-orange" onclick="resetDensity()">↺ Reset</button>
|
| 531 |
+
</div>
|
| 532 |
+
<div class="legend" style="margin-top:14px">
|
| 533 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--cyan)"></div>Original minority points</div>
|
| 534 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--green)"></div>✓ Accepted synthetic points</div>
|
| 535 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--red)"></div>✗ Rejected (outside d_mean)</div>
|
| 536 |
+
<div class="legend-item"><div class="legend-line" style="background:var(--cyan);border: 1px dashed var(--cyan)"></div>d_mean radius boundary</div>
|
| 537 |
+
</div>
|
| 538 |
+
<div class="insight">
|
| 539 |
+
<div class="insight-label">// The Innovation</div>
|
| 540 |
+
<p>A synthetic sample x_new is accepted only if ‖x_new − x_i‖ ≤ d_mean. This density gate keeps new points inside the safe zone — preventing noise, overfitting, and cluster bleeding.</p>
|
| 541 |
+
<div class="formula">if ‖x_new − xᵢ‖₂ ≤ d_mean → ACCEPT ✓ else → REJECT ✗</div>
|
| 542 |
+
</div>
|
| 543 |
+
</div>
|
| 544 |
+
</div>
|
| 545 |
+
|
| 546 |
+
<!-- ======================== PANEL 5: BEFORE / AFTER ======================== -->
|
| 547 |
+
<div class="panel" id="panel-4">
|
| 548 |
+
<div class="card">
|
| 549 |
+
<div class="card-title">Before vs After DSMOTE — KDD Cup 99</div>
|
| 550 |
+
<div class="card-sub">// Class distribution transformation after applying DSMOTE</div>
|
| 551 |
+
<div class="ba-row">
|
| 552 |
+
<div class="ba-card">
|
| 553 |
+
<h4 style="color:var(--orange)">⚠ Before — Severe Imbalance</h4>
|
| 554 |
+
<div id="before-bars"></div>
|
| 555 |
+
</div>
|
| 556 |
+
<div class="ba-card">
|
| 557 |
+
<h4 style="color:var(--green)">✓ After DSMOTE — Balanced</h4>
|
| 558 |
+
<div id="after-bars"></div>
|
| 559 |
+
</div>
|
| 560 |
+
</div>
|
| 561 |
+
<div style="margin-top:16px">
|
| 562 |
+
<button class="btn" onclick="animateBars()">▶ Animate Transformation</button>
|
| 563 |
+
<button class="btn btn-orange" onclick="resetBars()">↺ Reset</button>
|
| 564 |
+
</div>
|
| 565 |
+
<div class="insight">
|
| 566 |
+
<div class="insight-label">// Impact</div>
|
| 567 |
+
<p>Minority classes like <code style="color:var(--cyan)">pod</code> (264 samples) and <code style="color:var(--cyan)">warezclient</code> (1,020 samples) are boosted to ~256K–264K samples, achieving near-parity with the majority class after controlled reduction.</p>
|
| 568 |
+
</div>
|
| 569 |
+
</div>
|
| 570 |
+
|
| 571 |
+
<!-- F1 comparison -->
|
| 572 |
+
<div class="card">
|
| 573 |
+
<div class="card-title">Macro-F1 Score Comparison — UNSW-NF Dataset</div>
|
| 574 |
+
<div class="card-sub">// DSMOTE vs conventional oversampling methods</div>
|
| 575 |
+
<canvas id="f1-canvas" width="1040" height="280" style="width:100%"></canvas>
|
| 576 |
+
<button class="btn" style="margin-top:16px" onclick="animateF1()">▶ Show Results</button>
|
| 577 |
+
<div class="legend" style="margin-top:12px">
|
| 578 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--muted)"></div>ROS</div>
|
| 579 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--purple)"></div>SMOTE</div>
|
| 580 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--orange)"></div>Borderline-SMOTE</div>
|
| 581 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--yellow)"></div>ADASYN</div>
|
| 582 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--cyan)"></div>DSMOTE (Proposed)</div>
|
| 583 |
+
</div>
|
| 584 |
+
</div>
|
| 585 |
+
</div>
|
| 586 |
+
|
| 587 |
+
<!-- ======================== PANEL 6: UNSW RESULTS ======================== -->
|
| 588 |
+
<div class="panel" id="panel-5">
|
| 589 |
+
|
| 590 |
+
<!-- Stat strip -->
|
| 591 |
+
<div style="display:grid;grid-template-columns:repeat(4,1fr);gap:12px;margin-bottom:20px">
|
| 592 |
+
<div class="card" style="padding:16px;text-align:center;border-color:rgba(0,230,118,0.3)">
|
| 593 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted);letter-spacing:2px;margin-bottom:6px">DSMOTE BEST F1</div>
|
| 594 |
+
<div style="font-family:'Syne',sans-serif;font-size:1.8rem;font-weight:800;color:var(--green)">0.588</div>
|
| 595 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted)">RF Model</div>
|
| 596 |
+
</div>
|
| 597 |
+
<div class="card" style="padding:16px;text-align:center;border-color:rgba(255,23,68,0.3)">
|
| 598 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted);letter-spacing:2px;margin-bottom:6px">SMOTE BEST F1</div>
|
| 599 |
+
<div style="font-family:'Syne',sans-serif;font-size:1.8rem;font-weight:800;color:var(--red)">0.096</div>
|
| 600 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted)">RF / DT Model</div>
|
| 601 |
+
</div>
|
| 602 |
+
<div class="card" style="padding:16px;text-align:center;border-color:rgba(0,229,255,0.3)">
|
| 603 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted);letter-spacing:2px;margin-bottom:6px">IMPROVEMENT</div>
|
| 604 |
+
<div style="font-family:'Syne',sans-serif;font-size:1.8rem;font-weight:800;color:var(--cyan)">6.1×</div>
|
| 605 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted)">vs SMOTE RF</div>
|
| 606 |
+
</div>
|
| 607 |
+
<div class="card" style="padding:16px;text-align:center;border-color:rgba(255,215,64,0.3)">
|
| 608 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted);letter-spacing:2px;margin-bottom:6px">BAL. ACCURACY</div>
|
| 609 |
+
<div style="font-family:'Syne',sans-serif;font-size:1.8rem;font-weight:800;color:var(--yellow)">0.573</div>
|
| 610 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted)">DSMOTE RF</div>
|
| 611 |
+
</div>
|
| 612 |
+
</div>
|
| 613 |
+
|
| 614 |
+
<div class="card">
|
| 615 |
+
<div class="card-title">Macro-F1 by Model — All Methods (UNSW-NF)</div>
|
| 616 |
+
<div class="card-sub">// Real experimental results — DSMOTE is the ONLY method that meaningfully works on UNSW</div>
|
| 617 |
+
<div style="margin-bottom:12px;display:flex;gap:8px;flex-wrap:wrap" id="unsw-model-btns">
|
| 618 |
+
<button class="btn" style="padding:6px 12px" onclick="filterUnswModel('ALL')">All Models</button>
|
| 619 |
+
<button class="btn" style="padding:6px 12px" onclick="filterUnswModel('DT')">DT</button>
|
| 620 |
+
<button class="btn" style="padding:6px 12px" onclick="filterUnswModel('RF')">RF</button>
|
| 621 |
+
<button class="btn" style="padding:6px 12px" onclick="filterUnswModel('XGBoost')">XGBoost</button>
|
| 622 |
+
<button class="btn" style="padding:6px 12px" onclick="filterUnswModel('ANN')">ANN</button>
|
| 623 |
+
<button class="btn" style="padding:6px 12px" onclick="filterUnswModel('LSTM')">LSTM</button>
|
| 624 |
+
</div>
|
| 625 |
+
<canvas id="unsw-f1-canvas" width="1040" height="320" style="width:100%"></canvas>
|
| 626 |
+
<div class="legend" style="margin-top:12px">
|
| 627 |
+
<div class="legend-item"><div class="legend-dot" style="background:#78909c"></div>RAW</div>
|
| 628 |
+
<div class="legend-item"><div class="legend-dot" style="background:#546e7a"></div>ROS</div>
|
| 629 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--purple)"></div>SMOTE</div>
|
| 630 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--orange)"></div>BSMOTE</div>
|
| 631 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--yellow)"></div>ADASYN</div>
|
| 632 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--cyan)"></div>★ DSMOTE</div>
|
| 633 |
+
</div>
|
| 634 |
+
</div>
|
| 635 |
+
|
| 636 |
+
<div class="card">
|
| 637 |
+
<div class="card-title">Confusion Matrix Collapse — SMOTE vs DSMOTE (UNSW-NF, RF)</div>
|
| 638 |
+
<div class="card-sub">// SMOTE predicts EVERYTHING as "Benign" — DSMOTE correctly distributes predictions</div>
|
| 639 |
+
<div style="display:grid;grid-template-columns:1fr 1fr;gap:24px">
|
| 640 |
+
<div>
|
| 641 |
+
<canvas id="cm-smote-canvas" width="480" height="400"></canvas>
|
| 642 |
+
<div class="plot-label" style="color:var(--red);margin-top:8px">⚠ SMOTE — Total Collapse (F1: 0.096)</div>
|
| 643 |
+
</div>
|
| 644 |
+
<div>
|
| 645 |
+
<canvas id="cm-dsmote-canvas" width="480" height="400"></canvas>
|
| 646 |
+
<div class="plot-label" style="color:var(--green);margin-top:8px">✓ DSMOTE — Proper Distribution (F1: 0.588)</div>
|
| 647 |
+
</div>
|
| 648 |
+
</div>
|
| 649 |
+
<div class="insight" style="margin-top:16px">
|
| 650 |
+
<div class="insight-label">// The Collapse Problem</div>
|
| 651 |
+
<p>Under SMOTE, RF learns to predict every single sample as "Benign" (91.9% accuracy by doing nothing). DSMOTE forces the model to actually learn minority attack classes — Exploits, Fuzzers, Backdoor, Shellcode — that matter for security.</p>
|
| 652 |
+
</div>
|
| 653 |
+
<button class="btn" style="margin-top:12px" onclick="drawConfusionMatrices()">▶ Draw Matrices</button>
|
| 654 |
+
</div>
|
| 655 |
+
|
| 656 |
+
<div class="card">
|
| 657 |
+
<div class="card-title">Balanced Accuracy — DSMOTE vs All Methods (UNSW-NF)</div>
|
| 658 |
+
<div class="card-sub">// Balanced accuracy weights each class equally — exposes true minority class performance</div>
|
| 659 |
+
<canvas id="unsw-ba-canvas" width="1040" height="260" style="width:100%"></canvas>
|
| 660 |
+
<button class="btn" style="margin-top:12px" onclick="drawUnswBA()">▶ Show Balanced Accuracy</button>
|
| 661 |
+
</div>
|
| 662 |
+
</div>
|
| 663 |
+
|
| 664 |
+
<!-- ======================== PANEL 7: KDD RESULTS ======================== -->
|
| 665 |
+
<div class="panel" id="panel-6">
|
| 666 |
+
|
| 667 |
+
<!-- Stat strip -->
|
| 668 |
+
<div style="display:grid;grid-template-columns:repeat(4,1fr);gap:12px;margin-bottom:20px">
|
| 669 |
+
<div class="card" style="padding:16px;text-align:center;border-color:rgba(0,230,118,0.3)">
|
| 670 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted);letter-spacing:2px;margin-bottom:6px">DSMOTE RF F1</div>
|
| 671 |
+
<div style="font-family:'Syne',sans-serif;font-size:1.8rem;font-weight:800;color:var(--green)">0.9954</div>
|
| 672 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted)">Matches RAW best</div>
|
| 673 |
+
</div>
|
| 674 |
+
<div class="card" style="padding:16px;text-align:center;border-color:rgba(0,229,255,0.3)">
|
| 675 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted);letter-spacing:2px;margin-bottom:6px">DSMOTE ANN F1</div>
|
| 676 |
+
<div style="font-family:'Syne',sans-serif;font-size:1.8rem;font-weight:800;color:var(--cyan)">0.9285</div>
|
| 677 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted)">+0.013 vs SMOTE ANN</div>
|
| 678 |
+
</div>
|
| 679 |
+
<div class="card" style="padding:16px;text-align:center;border-color:rgba(255,215,64,0.3)">
|
| 680 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted);letter-spacing:2px;margin-bottom:6px">BAL. ACCURACY</div>
|
| 681 |
+
<div style="font-family:'Syne',sans-serif;font-size:1.8rem;font-weight:800;color:var(--yellow)">0.9940</div>
|
| 682 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted)">DSMOTE RF</div>
|
| 683 |
+
</div>
|
| 684 |
+
<div class="card" style="padding:16px;text-align:center;border-color:rgba(124,77,255,0.3)">
|
| 685 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted);letter-spacing:2px;margin-bottom:6px">G-MEAN</div>
|
| 686 |
+
<div style="font-family:'Syne',sans-serif;font-size:1.8rem;font-weight:800;color:var(--purple)">0.9939</div>
|
| 687 |
+
<div style="font-family:'JetBrains Mono',monospace;font-size:9px;color:var(--muted)">DSMOTE RF</div>
|
| 688 |
+
</div>
|
| 689 |
+
</div>
|
| 690 |
+
|
| 691 |
+
<div class="card">
|
| 692 |
+
<div class="card-title">Macro-F1 by Model — All Methods (KDD Cup 99)</div>
|
| 693 |
+
<div class="card-sub">// KDD is harder to fail on — DSMOTE matches top performance while improving minority class stability</div>
|
| 694 |
+
<canvas id="kdd-f1-canvas" width="1040" height="320" style="width:100%"></canvas>
|
| 695 |
+
<div class="legend" style="margin-top:12px">
|
| 696 |
+
<div class="legend-item"><div class="legend-dot" style="background:#78909c"></div>RAW</div>
|
| 697 |
+
<div class="legend-item"><div class="legend-dot" style="background:#546e7a"></div>ROS</div>
|
| 698 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--purple)"></div>SMOTE</div>
|
| 699 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--orange)"></div>BSMOTE</div>
|
| 700 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--cyan)"></div>★ DSMOTE</div>
|
| 701 |
+
</div>
|
| 702 |
+
</div>
|
| 703 |
+
|
| 704 |
+
<div class="card">
|
| 705 |
+
<div class="card-title">Multi-Metric Radar — Best Models Comparison (KDD)</div>
|
| 706 |
+
<div class="card-sub">// DSMOTE (RF) vs RAW (RF): Accuracy · Balanced Acc · F1 · G-Mean · Precision · Recall</div>
|
| 707 |
+
<canvas id="kdd-radar-canvas" width="1040" height="360" style="width:100%"></canvas>
|
| 708 |
+
<button class="btn" style="margin-top:12px" onclick="drawRadar()">▶ Draw Radar</button>
|
| 709 |
+
<div class="legend" style="margin-top:12px">
|
| 710 |
+
<div class="legend-item"><div class="legend-dot" style="background:#78909c"></div>RAW RF</div>
|
| 711 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--cyan)"></div>DSMOTE RF</div>
|
| 712 |
+
<div class="legend-item"><div class="legend-dot" style="background:var(--purple)"></div>SMOTE RF</div>
|
| 713 |
+
</div>
|
| 714 |
+
</div>
|
| 715 |
+
|
| 716 |
+
<div class="card">
|
| 717 |
+
<div class="card-title">KDD — G-Mean Comparison Across Models</div>
|
| 718 |
+
<div class="card-sub">// G-Mean = geometric mean of per-class recalls — punishes models that ignore minority classes</div>
|
| 719 |
+
<canvas id="kdd-gmean-canvas" width="1040" height="260" style="width:100%"></canvas>
|
| 720 |
+
<button class="btn" style="margin-top:12px" onclick="drawKddGmean()">▶ Show G-Mean</button>
|
| 721 |
+
</div>
|
| 722 |
+
</div>
|
| 723 |
+
|
| 724 |
+
</div><!-- /container -->
|
| 725 |
+
|
| 726 |
+
<script>
|
| 727 |
+
// ============================================================
|
| 728 |
+
// UTILITIES
|
| 729 |
+
// ============================================================
|
| 730 |
+
function seededRng(seed) {
|
| 731 |
+
let s = seed;
|
| 732 |
+
return function() {
|
| 733 |
+
s = (s * 1664525 + 1013904223) & 0xffffffff;
|
| 734 |
+
return (s >>> 0) / 0xffffffff;
|
| 735 |
+
};
|
| 736 |
+
}
|
| 737 |
+
|
| 738 |
+
function gaussianPair(rng, mx, my, sx, sy) {
|
| 739 |
+
// Box-Muller
|
| 740 |
+
const u1 = rng(), u2 = rng();
|
| 741 |
+
const z0 = Math.sqrt(-2 * Math.log(u1 + 0.0001)) * Math.cos(2 * Math.PI * u2);
|
| 742 |
+
const z1 = Math.sqrt(-2 * Math.log(u1 + 0.0001)) * Math.sin(2 * Math.PI * u2);
|
| 743 |
+
return [mx + z0 * sx, my + z1 * sy];
|
| 744 |
+
}
|
| 745 |
+
|
| 746 |
+
function dataToCanvas(x, y, xmin, xmax, ymin, ymax, W, H, pad) {
|
| 747 |
+
const cx = pad + (x - xmin) / (xmax - xmin) * (W - 2 * pad);
|
| 748 |
+
const cy = H - pad - (y - ymin) / (ymax - ymin) * (H - 2 * pad);
|
| 749 |
+
return [cx, cy];
|
| 750 |
+
}
|
| 751 |
+
|
| 752 |
+
const PAD = 40;
|
| 753 |
+
|
| 754 |
+
function drawAxes(ctx, W, H) {
|
| 755 |
+
ctx.strokeStyle = 'rgba(0,229,255,0.12)';
|
| 756 |
+
ctx.lineWidth = 1;
|
| 757 |
+
// grid lines
|
| 758 |
+
for (let i = 0; i <= 5; i++) {
|
| 759 |
+
const x = PAD + i * (W - 2 * PAD) / 5;
|
| 760 |
+
const y = PAD + i * (H - 2 * PAD) / 5;
|
| 761 |
+
ctx.beginPath(); ctx.moveTo(x, PAD); ctx.lineTo(x, H - PAD); ctx.stroke();
|
| 762 |
+
ctx.beginPath(); ctx.moveTo(PAD, y); ctx.lineTo(W - PAD, y); ctx.stroke();
|
| 763 |
+
}
|
| 764 |
+
// axes
|
| 765 |
+
ctx.strokeStyle = 'rgba(0,229,255,0.25)';
|
| 766 |
+
ctx.beginPath(); ctx.moveTo(PAD, H - PAD); ctx.lineTo(W - PAD, H - PAD); ctx.stroke();
|
| 767 |
+
ctx.beginPath(); ctx.moveTo(PAD, PAD); ctx.lineTo(PAD, H - PAD); ctx.stroke();
|
| 768 |
+
// labels
|
| 769 |
+
ctx.fillStyle = 'rgba(84,110,122,0.8)';
|
| 770 |
+
ctx.font = '10px JetBrains Mono, monospace';
|
| 771 |
+
ctx.fillText('PC1 →', W - PAD - 2, H - PAD + 16);
|
| 772 |
+
ctx.save(); ctx.translate(PAD - 16, PAD + 10); ctx.rotate(-Math.PI / 2);
|
| 773 |
+
ctx.fillText('PC2 →', 0, 0); ctx.restore();
|
| 774 |
+
}
|
| 775 |
+
|
| 776 |
+
function dot(ctx, cx, cy, r, color, alpha = 1) {
|
| 777 |
+
ctx.globalAlpha = alpha;
|
| 778 |
+
ctx.fillStyle = color;
|
| 779 |
+
ctx.beginPath();
|
| 780 |
+
ctx.arc(cx, cy, r, 0, Math.PI * 2);
|
| 781 |
+
ctx.fill();
|
| 782 |
+
ctx.globalAlpha = 1;
|
| 783 |
+
}
|
| 784 |
+
|
| 785 |
+
// ============================================================
|
| 786 |
+
// TAB SWITCHING — handled at bottom of script
|
| 787 |
+
// ============================================================
|
| 788 |
+
|
| 789 |
+
// ============================================================
|
| 790 |
+
// PANEL 1: SMOTE vs DSMOTE
|
| 791 |
+
// ============================================================
|
| 792 |
+
const rng1 = seededRng(42);
|
| 793 |
+
const XMIN = -5, XMAX = 5, YMIN = -5, YMAX = 5;
|
| 794 |
+
|
| 795 |
+
function genMinorityClusters() {
|
| 796 |
+
const r = seededRng(7);
|
| 797 |
+
const pts = [];
|
| 798 |
+
// Cluster A: top-left
|
| 799 |
+
for (let i = 0; i < 30; i++) {
|
| 800 |
+
const [x, y] = gaussianPair(r, -2.8, 2.2, 0.6, 0.6);
|
| 801 |
+
pts.push({ x, y, cluster: 0 });
|
| 802 |
+
}
|
| 803 |
+
// Cluster B: bottom-right
|
| 804 |
+
for (let i = 0; i < 25; i++) {
|
| 805 |
+
const [x, y] = gaussianPair(r, 2.8, -2.2, 0.55, 0.55);
|
| 806 |
+
pts.push({ x, y, cluster: 1 });
|
| 807 |
+
}
|
| 808 |
+
// Cluster C: middle-right (smaller)
|
| 809 |
+
for (let i = 0; i < 18; i++) {
|
| 810 |
+
const [x, y] = gaussianPair(r, 1.2, 1.8, 0.4, 0.4);
|
| 811 |
+
pts.push({ x, y, cluster: 2 });
|
| 812 |
+
}
|
| 813 |
+
return pts;
|
| 814 |
+
}
|
| 815 |
+
|
| 816 |
+
function genMajority() {
|
| 817 |
+
const r = seededRng(11);
|
| 818 |
+
const pts = [];
|
| 819 |
+
for (let i = 0; i < 200; i++) {
|
| 820 |
+
const x = (r() - 0.5) * 8;
|
| 821 |
+
const y = (r() - 0.5) * 8;
|
| 822 |
+
pts.push({ x, y });
|
| 823 |
+
}
|
| 824 |
+
return pts;
|
| 825 |
+
}
|
| 826 |
+
|
| 827 |
+
const minPts = genMinorityClusters();
|
| 828 |
+
const majPts = genMajority();
|
| 829 |
+
let smoteAnimPts = [], dsmoteAnimPts = [];
|
| 830 |
+
let smoteAnim = null, dsmoteAnim = null;
|
| 831 |
+
|
| 832 |
+
function drawScatterBase(canvasId, synthPts, synthColor, label) {
|
| 833 |
+
const canvas = document.getElementById(canvasId);
|
| 834 |
+
const ctx = canvas.getContext('2d');
|
| 835 |
+
const W = canvas.width, H = canvas.height;
|
| 836 |
+
ctx.clearRect(0, 0, W, H);
|
| 837 |
+
drawAxes(ctx, W, H);
|
| 838 |
+
|
| 839 |
+
// majority
|
| 840 |
+
majPts.forEach(p => {
|
| 841 |
+
const [cx, cy] = dataToCanvas(p.x, p.y, XMIN, XMAX, YMIN, YMAX, W, H, PAD);
|
| 842 |
+
dot(ctx, cx, cy, 3, '#546e7a', 0.3);
|
| 843 |
+
});
|
| 844 |
+
|
| 845 |
+
// minority originals
|
| 846 |
+
const clColors = ['#00e5ff', '#7c4dff', '#ffd740'];
|
| 847 |
+
minPts.forEach(p => {
|
| 848 |
+
const [cx, cy] = dataToCanvas(p.x, p.y, XMIN, XMAX, YMIN, YMAX, W, H, PAD);
|
| 849 |
+
dot(ctx, cx, cy, 5, clColors[p.cluster]);
|
| 850 |
+
});
|
| 851 |
+
|
| 852 |
+
// synthetic
|
| 853 |
+
synthPts.forEach(p => {
|
| 854 |
+
const [cx, cy] = dataToCanvas(p.x, p.y, XMIN, XMAX, YMIN, YMAX, W, H, PAD);
|
| 855 |
+
dot(ctx, cx, cy, 4, synthColor, 0.75);
|
| 856 |
+
});
|
| 857 |
+
}
|
| 858 |
+
|
| 859 |
+
function generateSMOTEPoint() {
|
| 860 |
+
const r = seededRng(smoteAnimPts.length * 13 + 7);
|
| 861 |
+
// pick ANY two minority pts (ignoring cluster)
|
| 862 |
+
const i = Math.floor(r() * minPts.length);
|
| 863 |
+
const j = Math.floor(r() * minPts.length);
|
| 864 |
+
const lam = r();
|
| 865 |
+
return {
|
| 866 |
+
x: minPts[i].x + lam * (minPts[j].x - minPts[i].x),
|
| 867 |
+
y: minPts[i].y + lam * (minPts[j].y - minPts[i].y)
|
| 868 |
+
};
|
| 869 |
+
}
|
| 870 |
+
|
| 871 |
+
function generateDSMOTEPoint(idx) {
|
| 872 |
+
const r = seededRng(idx * 17 + 3);
|
| 873 |
+
// Pick a cluster
|
| 874 |
+
const clusterPts = [
|
| 875 |
+
minPts.filter(p => p.cluster === 0),
|
| 876 |
+
minPts.filter(p => p.cluster === 1),
|
| 877 |
+
minPts.filter(p => p.cluster === 2)
|
| 878 |
+
];
|
| 879 |
+
const cl = Math.floor(r() * 3);
|
| 880 |
+
const pts = clusterPts[cl];
|
| 881 |
+
const i = Math.floor(r() * pts.length);
|
| 882 |
+
const j = Math.floor(r() * pts.length);
|
| 883 |
+
const lam = r();
|
| 884 |
+
const nx = pts[i].x + lam * (pts[j].x - pts[i].x);
|
| 885 |
+
const ny = pts[i].y + lam * (pts[j].y - pts[i].y);
|
| 886 |
+
// density check: within cluster
|
| 887 |
+
const cx_mean = pts.reduce((a, p) => a + p.x, 0) / pts.length;
|
| 888 |
+
const cy_mean = pts.reduce((a, p) => a + p.y, 0) / pts.length;
|
| 889 |
+
const d = Math.sqrt((nx - pts[i].x) ** 2 + (ny - pts[i].y) ** 2);
|
| 890 |
+
// compute mean intra dist (simplified)
|
| 891 |
+
const dmean = 0.9;
|
| 892 |
+
if (d <= dmean) return { x: nx, y: ny };
|
| 893 |
+
return null;
|
| 894 |
+
}
|
| 895 |
+
|
| 896 |
+
function animateSMOTE() {
|
| 897 |
+
if (smoteAnim) clearInterval(smoteAnim);
|
| 898 |
+
smoteAnimPts = [];
|
| 899 |
+
let count = 0;
|
| 900 |
+
smoteAnim = setInterval(() => {
|
| 901 |
+
if (count >= 80) { clearInterval(smoteAnim); return; }
|
| 902 |
+
smoteAnimPts.push(generateSMOTEPoint());
|
| 903 |
+
drawScatterBase('smote-canvas', smoteAnimPts, '#ff6b35', 'SMOTE');
|
| 904 |
+
count++;
|
| 905 |
+
}, 40);
|
| 906 |
+
}
|
| 907 |
+
|
| 908 |
+
function animateDSMOTE() {
|
| 909 |
+
if (dsmoteAnim) clearInterval(dsmoteAnim);
|
| 910 |
+
dsmoteAnimPts = [];
|
| 911 |
+
let count = 0;
|
| 912 |
+
dsmoteAnim = setInterval(() => {
|
| 913 |
+
if (count >= 80) { clearInterval(dsmoteAnim); return; }
|
| 914 |
+
let p = null, tries = 0;
|
| 915 |
+
while (!p && tries < 20) { p = generateDSMOTEPoint(count * 7 + tries); tries++; }
|
| 916 |
+
if (p) dsmoteAnimPts.push(p);
|
| 917 |
+
drawScatterBase('dsmote-canvas', dsmoteAnimPts, '#00e676', 'DSMOTE');
|
| 918 |
+
count++;
|
| 919 |
+
}, 40);
|
| 920 |
+
}
|
| 921 |
+
|
| 922 |
+
function resetSmote() {
|
| 923 |
+
if (smoteAnim) clearInterval(smoteAnim);
|
| 924 |
+
if (dsmoteAnim) clearInterval(dsmoteAnim);
|
| 925 |
+
smoteAnimPts = []; dsmoteAnimPts = [];
|
| 926 |
+
drawScatterBase('smote-canvas', [], '#ff6b35');
|
| 927 |
+
drawScatterBase('dsmote-canvas', [], '#00e676');
|
| 928 |
+
}
|
| 929 |
+
|
| 930 |
+
// ============================================================
|
| 931 |
+
// PANEL 2: PIPELINE
|
| 932 |
+
// ============================================================
|
| 933 |
+
const stepDetails = [
|
| 934 |
+
{
|
| 935 |
+
color: '#ff6b35', title: 'Step 1 — Majority Class Reduction',
|
| 936 |
+
desc: 'The dominant class (e.g., smurf with 2.8M samples) is randomly down-sampled by keeping only a fraction p of its data. This reduces bias in training without losing minority class information.',
|
| 937 |
+
formula: 'X_maj_reduced = X_maj[0 : ρ × N_maj] (ρ = 0.5)'
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
color: '#7c4dff', title: 'Step 2 — PCA Dimensionality Reduction',
|
| 941 |
+
desc: 'Principal Component Analysis reduces the feature space while retaining 95% of the variance. This makes KMeans clustering and KNN more effective and computationally efficient.',
|
| 942 |
+
formula: 'X_pca = PCA(X, variance_ratio = 0.95)'
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
color: '#00e5ff', title: 'Step 3 — KMeans Clustering',
|
| 946 |
+
desc: 'For each minority class, KMeans groups the samples into k sub-clusters. This lets DSMOTE understand the internal structure of each attack type rather than treating them as a uniform blob.',
|
| 947 |
+
formula: 'cluster_labels = KMeans(X_Cm, k=3)'
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
color: '#ffd740', title: 'Step 4 — KNN + Interpolation',
|
| 951 |
+
desc: 'Within each cluster, K-nearest neighbors are computed. Synthetic samples are generated by interpolating between a selected point and one of its neighbors — just like SMOTE, but cluster-confined.',
|
| 952 |
+
formula: 'x_new = xᵢ + λ × (x_j − xᵢ), λ ~ U(0,1)'
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
color: '#00e676', title: 'Step 5 — Density Constraint Filter',
|
| 956 |
+
desc: 'A synthetic sample is accepted only if it falls within the mean intra-cluster distance. This prevents points from being generated in sparse, noisy regions outside the true data distribution.',
|
| 957 |
+
formula: '‖x_new − xᵢ‖₂ ≤ d_mean → ACCEPT'
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
color: '#ff4081', title: 'Step 6 — Class Weight Computation',
|
| 961 |
+
desc: 'After oversampling, class weights are computed inversely proportional to class frequency. These weights guide the loss function during training to further emphasize minority classes.',
|
| 962 |
+
formula: 'w_c = N_total / (K × N_c)'
|
| 963 |
+
}
|
| 964 |
+
];
|
| 965 |
+
|
| 966 |
+
function showStep(i) {
|
| 967 |
+
const detail = document.getElementById('step-detail');
|
| 968 |
+
const s = stepDetails[i];
|
| 969 |
+
detail.className = 'step-detail visible';
|
| 970 |
+
detail.style.borderLeftColor = s.color;
|
| 971 |
+
detail.innerHTML = `
|
| 972 |
+
<h4 style="color:${s.color}">${s.title}</h4>
|
| 973 |
+
<p>${s.desc}</p>
|
| 974 |
+
<div class="formula" style="color:${s.color}">${s.formula}</div>
|
| 975 |
+
`;
|
| 976 |
+
document.querySelectorAll('.pipe-icon').forEach((el, j) => {
|
| 977 |
+
el.classList.toggle('active-step', i === j);
|
| 978 |
+
});
|
| 979 |
+
}
|
| 980 |
+
|
| 981 |
+
function initPipeline() {
|
| 982 |
+
document.querySelectorAll('.pipe-step').forEach((el, i) => {
|
| 983 |
+
setTimeout(() => el.classList.add('visible'), i * 120);
|
| 984 |
+
});
|
| 985 |
+
drawPipelineCanvas();
|
| 986 |
+
}
|
| 987 |
+
|
| 988 |
+
function drawPipelineCanvas() {
|
| 989 |
+
const canvas = document.getElementById('pipeline-canvas');
|
| 990 |
+
if (!canvas) return;
|
| 991 |
+
const ctx = canvas.getContext('2d');
|
| 992 |
+
const W = canvas.width, H = canvas.height;
|
| 993 |
+
ctx.clearRect(0, 0, W, H);
|
| 994 |
+
|
| 995 |
+
const steps = [
|
| 996 |
+
{ label: 'RAW DATA', color: '#546e7a', val: '4.9M rows\n41 features' },
|
| 997 |
+
{ label: 'MAJORITY ↓', color: '#ff6b35', val: '2.45M rows\n41 features' },
|
| 998 |
+
{ label: 'PCA', color: '#7c4dff', val: '2.45M rows\n~18 PCs' },
|
| 999 |
+
{ label: 'CLUSTERING', color: '#00e5ff', val: 'k clusters\nper class' },
|
| 1000 |
+
{ label: 'SYNTHETIC', color: '#ffd740', val: '+1.8M\nnew samples' },
|
| 1001 |
+
{ label: 'BALANCED', color: '#00e676', val: '~270K\nper class' },
|
| 1002 |
+
];
|
| 1003 |
+
|
| 1004 |
+
const bw = 120, bh = 80, gap = (W - steps.length * bw) / (steps.length + 1);
|
| 1005 |
+
steps.forEach((s, i) => {
|
| 1006 |
+
const x = gap + i * (bw + gap);
|
| 1007 |
+
const y = (H - bh) / 2;
|
| 1008 |
+
// box
|
| 1009 |
+
ctx.strokeStyle = s.color;
|
| 1010 |
+
ctx.lineWidth = 1.5;
|
| 1011 |
+
ctx.strokeRect(x, y, bw, bh);
|
| 1012 |
+
ctx.fillStyle = s.color + '15';
|
| 1013 |
+
ctx.fillRect(x, y, bw, bh);
|
| 1014 |
+
// label
|
| 1015 |
+
ctx.fillStyle = s.color;
|
| 1016 |
+
ctx.font = 'bold 11px JetBrains Mono, monospace';
|
| 1017 |
+
ctx.textAlign = 'center';
|
| 1018 |
+
ctx.fillText(s.label, x + bw / 2, y + 22);
|
| 1019 |
+
// value
|
| 1020 |
+
ctx.fillStyle = 'rgba(224,247,250,0.55)';
|
| 1021 |
+
ctx.font = '10px JetBrains Mono, monospace';
|
| 1022 |
+
const lines = s.val.split('\n');
|
| 1023 |
+
lines.forEach((l, li) => ctx.fillText(l, x + bw / 2, y + 42 + li * 15));
|
| 1024 |
+
// arrow
|
| 1025 |
+
if (i < steps.length - 1) {
|
| 1026 |
+
const ax = x + bw + 4, ay = H / 2;
|
| 1027 |
+
ctx.strokeStyle = 'rgba(84,110,122,0.6)';
|
| 1028 |
+
ctx.lineWidth = 1;
|
| 1029 |
+
ctx.beginPath(); ctx.moveTo(ax, ay); ctx.lineTo(ax + gap - 8, ay); ctx.stroke();
|
| 1030 |
+
ctx.fillStyle = 'rgba(84,110,122,0.6)';
|
| 1031 |
+
ctx.beginPath();
|
| 1032 |
+
ctx.moveTo(ax + gap - 8, ay - 5);
|
| 1033 |
+
ctx.lineTo(ax + gap - 1, ay);
|
| 1034 |
+
ctx.lineTo(ax + gap - 8, ay + 5);
|
| 1035 |
+
ctx.fill();
|
| 1036 |
+
}
|
| 1037 |
+
});
|
| 1038 |
+
ctx.textAlign = 'left';
|
| 1039 |
+
}
|
| 1040 |
+
|
| 1041 |
+
// ============================================================
|
| 1042 |
+
// PANEL 3: CLUSTERING
|
| 1043 |
+
// ============================================================
|
| 1044 |
+
let clusterAnimDone = false;
|
| 1045 |
+
const clRng = seededRng(55);
|
| 1046 |
+
|
| 1047 |
+
function genClusterPts() {
|
| 1048 |
+
const centers = [[-2, 2.5], [2.5, 0.5], [-0.5, -2.5], [3.5, -2.5]];
|
| 1049 |
+
const colors = ['#00e5ff', '#ff6b35', '#7c4dff', '#ffd740'];
|
| 1050 |
+
const pts = [];
|
| 1051 |
+
centers.forEach((c, ci) => {
|
| 1052 |
+
const n = 22 + Math.floor(clRng() * 10);
|
| 1053 |
+
for (let i = 0; i < n; i++) {
|
| 1054 |
+
const [x, y] = gaussianPair(clRng, c[0], c[1], 0.55, 0.55);
|
| 1055 |
+
pts.push({ x, y, cluster: ci, color: colors[ci] });
|
| 1056 |
+
}
|
| 1057 |
+
});
|
| 1058 |
+
return pts;
|
| 1059 |
+
}
|
| 1060 |
+
|
| 1061 |
+
const clusterPts = genClusterPts();
|
| 1062 |
+
let clustersRevealed = false;
|
| 1063 |
+
|
| 1064 |
+
function drawClusters(revealed) {
|
| 1065 |
+
const canvas = document.getElementById('cluster-canvas');
|
| 1066 |
+
if (!canvas) return;
|
| 1067 |
+
const ctx = canvas.getContext('2d');
|
| 1068 |
+
const W = canvas.width, H = canvas.height;
|
| 1069 |
+
ctx.clearRect(0, 0, W, H);
|
| 1070 |
+
drawAxes(ctx, W, H);
|
| 1071 |
+
|
| 1072 |
+
clusterPts.forEach(p => {
|
| 1073 |
+
const [cx, cy] = dataToCanvas(p.x, p.y, -5, 5, -5, 5, W, H, PAD);
|
| 1074 |
+
const col = revealed ? p.color : '#546e7a';
|
| 1075 |
+
dot(ctx, cx, cy, 5.5, col, revealed ? 0.85 : 0.5);
|
| 1076 |
+
});
|
| 1077 |
+
|
| 1078 |
+
if (revealed) {
|
| 1079 |
+
// draw centroid markers
|
| 1080 |
+
const centers = [[-2, 2.5], [2.5, 0.5], [-0.5, -2.5], [3.5, -2.5]];
|
| 1081 |
+
const colors = ['#00e5ff', '#ff6b35', '#7c4dff', '#ffd740'];
|
| 1082 |
+
centers.forEach((c, ci) => {
|
| 1083 |
+
const [cx, cy] = dataToCanvas(c[0], c[1], -5, 5, -5, 5, W, H, PAD);
|
| 1084 |
+
ctx.strokeStyle = colors[ci];
|
| 1085 |
+
ctx.lineWidth = 2;
|
| 1086 |
+
ctx.beginPath();
|
| 1087 |
+
ctx.moveTo(cx - 8, cy); ctx.lineTo(cx + 8, cy);
|
| 1088 |
+
ctx.moveTo(cx, cy - 8); ctx.lineTo(cx, cy + 8);
|
| 1089 |
+
ctx.stroke();
|
| 1090 |
+
ctx.font = 'bold 10px JetBrains Mono, monospace';
|
| 1091 |
+
ctx.fillStyle = colors[ci];
|
| 1092 |
+
ctx.fillText(`C${ci + 1}`, cx + 10, cy - 6);
|
| 1093 |
+
});
|
| 1094 |
+
|
| 1095 |
+
// label
|
| 1096 |
+
ctx.font = 'bold 13px Rajdhani, sans-serif';
|
| 1097 |
+
ctx.fillStyle = 'rgba(0,229,255,0.7)';
|
| 1098 |
+
ctx.fillText('✓ KMeans reveals sub-structure', PAD + 5, PAD + 18);
|
| 1099 |
+
} else {
|
| 1100 |
+
ctx.font = 'bold 13px Rajdhani, sans-serif';
|
| 1101 |
+
ctx.fillStyle = 'rgba(84,110,122,0.7)';
|
| 1102 |
+
ctx.fillText('Minority samples (unclustered)', PAD + 5, PAD + 18);
|
| 1103 |
+
}
|
| 1104 |
+
}
|
| 1105 |
+
|
| 1106 |
+
function initClusters() { drawClusters(false); }
|
| 1107 |
+
function animateClusters() {
|
| 1108 |
+
drawClusters(false);
|
| 1109 |
+
setTimeout(() => drawClusters(true), 600);
|
| 1110 |
+
}
|
| 1111 |
+
function resetClusters() { drawClusters(false); }
|
| 1112 |
+
|
| 1113 |
+
// ============================================================
|
| 1114 |
+
// PANEL 4: DENSITY CONSTRAINT
|
| 1115 |
+
// ============================================================
|
| 1116 |
+
let densityPts = [], densityAnim = null;
|
| 1117 |
+
const dRng = seededRng(88);
|
| 1118 |
+
|
| 1119 |
+
function genDensitySourcePts() {
|
| 1120 |
+
const pts = [];
|
| 1121 |
+
for (let i = 0; i < 25; i++) {
|
| 1122 |
+
const [x, y] = gaussianPair(dRng, 0, 0, 0.9, 0.9);
|
| 1123 |
+
pts.push({ x, y });
|
| 1124 |
+
}
|
| 1125 |
+
return pts;
|
| 1126 |
+
}
|
| 1127 |
+
const densitySrcPts = genDensitySourcePts();
|
| 1128 |
+
const D_MEAN = 1.1; // in data units
|
| 1129 |
+
|
| 1130 |
+
function computeDMean(pts) {
|
| 1131 |
+
let total = 0, count = 0;
|
| 1132 |
+
pts.forEach(p => {
|
| 1133 |
+
pts.forEach(q => {
|
| 1134 |
+
total += Math.sqrt((p.x - q.x) ** 2 + (p.y - q.y) ** 2);
|
| 1135 |
+
count++;
|
| 1136 |
+
});
|
| 1137 |
+
});
|
| 1138 |
+
return total / count;
|
| 1139 |
+
}
|
| 1140 |
+
|
| 1141 |
+
function drawDensity(accepted, rejected) {
|
| 1142 |
+
const canvas = document.getElementById('density-canvas');
|
| 1143 |
+
if (!canvas) return;
|
| 1144 |
+
const ctx = canvas.getContext('2d');
|
| 1145 |
+
const W = canvas.width, H = canvas.height;
|
| 1146 |
+
ctx.clearRect(0, 0, W, H);
|
| 1147 |
+
drawAxes(ctx, W, H);
|
| 1148 |
+
|
| 1149 |
+
// draw d_mean circles for each original point
|
| 1150 |
+
densitySrcPts.forEach(p => {
|
| 1151 |
+
const [cx, cy] = dataToCanvas(p.x, p.y, -5, 5, -5, 5, W, H, PAD);
|
| 1152 |
+
const r_px = D_MEAN / 10 * (W - 2 * PAD);
|
| 1153 |
+
ctx.strokeStyle = 'rgba(0,229,255,0.12)';
|
| 1154 |
+
ctx.lineWidth = 1;
|
| 1155 |
+
ctx.setLineDash([4, 4]);
|
| 1156 |
+
ctx.beginPath(); ctx.arc(cx, cy, r_px, 0, Math.PI * 2); ctx.stroke();
|
| 1157 |
+
ctx.setLineDash([]);
|
| 1158 |
+
});
|
| 1159 |
+
|
| 1160 |
+
// draw one prominent circle
|
| 1161 |
+
const refPt = densitySrcPts[5];
|
| 1162 |
+
const [rcx, rcy] = dataToCanvas(refPt.x, refPt.y, -5, 5, -5, 5, W, H, PAD);
|
| 1163 |
+
const r_px = D_MEAN / 10 * (W - 2 * PAD);
|
| 1164 |
+
ctx.strokeStyle = 'rgba(0,229,255,0.55)';
|
| 1165 |
+
ctx.lineWidth = 2;
|
| 1166 |
+
ctx.setLineDash([6, 4]);
|
| 1167 |
+
ctx.beginPath(); ctx.arc(rcx, rcy, r_px, 0, Math.PI * 2); ctx.stroke();
|
| 1168 |
+
ctx.setLineDash([]);
|
| 1169 |
+
|
| 1170 |
+
// label
|
| 1171 |
+
ctx.fillStyle = 'rgba(0,229,255,0.5)';
|
| 1172 |
+
ctx.font = '10px JetBrains Mono, monospace';
|
| 1173 |
+
ctx.fillText('d_mean', rcx + r_px + 5, rcy - 5);
|
| 1174 |
+
|
| 1175 |
+
// original pts
|
| 1176 |
+
densitySrcPts.forEach(p => {
|
| 1177 |
+
const [cx, cy] = dataToCanvas(p.x, p.y, -5, 5, -5, 5, W, H, PAD);
|
| 1178 |
+
dot(ctx, cx, cy, 5.5, '#00e5ff');
|
| 1179 |
+
});
|
| 1180 |
+
|
| 1181 |
+
// rejected
|
| 1182 |
+
rejected.forEach(p => {
|
| 1183 |
+
const [cx, cy] = dataToCanvas(p.x, p.y, -5, 5, -5, 5, W, H, PAD);
|
| 1184 |
+
dot(ctx, cx, cy, 4.5, '#ff1744', 0.75);
|
| 1185 |
+
// X mark
|
| 1186 |
+
ctx.strokeStyle = '#ff1744';
|
| 1187 |
+
ctx.lineWidth = 1.5;
|
| 1188 |
+
ctx.beginPath(); ctx.moveTo(cx - 4, cy - 4); ctx.lineTo(cx + 4, cy + 4); ctx.stroke();
|
| 1189 |
+
ctx.beginPath(); ctx.moveTo(cx + 4, cy - 4); ctx.lineTo(cx - 4, cy + 4); ctx.stroke();
|
| 1190 |
+
});
|
| 1191 |
+
|
| 1192 |
+
// accepted
|
| 1193 |
+
accepted.forEach(p => {
|
| 1194 |
+
const [cx, cy] = dataToCanvas(p.x, p.y, -5, 5, -5, 5, W, H, PAD);
|
| 1195 |
+
dot(ctx, cx, cy, 4.5, '#00e676', 0.8);
|
| 1196 |
+
});
|
| 1197 |
+
|
| 1198 |
+
// counts
|
| 1199 |
+
ctx.font = 'bold 12px JetBrains Mono, monospace';
|
| 1200 |
+
ctx.fillStyle = '#00e676';
|
| 1201 |
+
ctx.fillText(`✓ Accepted: ${accepted.length}`, PAD + 5, PAD + 16);
|
| 1202 |
+
ctx.fillStyle = '#ff1744';
|
| 1203 |
+
ctx.fillText(`✗ Rejected: ${rejected.length}`, PAD + 5, PAD + 32);
|
| 1204 |
+
}
|
| 1205 |
+
|
| 1206 |
+
function genSyntheticDensityPt(idx) {
|
| 1207 |
+
const r = seededRng(idx * 31 + 17);
|
| 1208 |
+
const i = Math.floor(r() * densitySrcPts.length);
|
| 1209 |
+
const j = Math.floor(r() * densitySrcPts.length);
|
| 1210 |
+
const lam = r();
|
| 1211 |
+
const nx = densitySrcPts[i].x + lam * (densitySrcPts[j].x - densitySrcPts[i].x);
|
| 1212 |
+
const ny = densitySrcPts[i].y + lam * (densitySrcPts[j].y - densitySrcPts[i].y);
|
| 1213 |
+
const dist = Math.sqrt((nx - densitySrcPts[i].x) ** 2 + (ny - densitySrcPts[i].y) ** 2);
|
| 1214 |
+
return { x: nx, y: ny, accepted: dist <= D_MEAN };
|
| 1215 |
+
}
|
| 1216 |
+
|
| 1217 |
+
let densityAccepted = [], densityRejected = [];
|
| 1218 |
+
|
| 1219 |
+
function initDensity() { drawDensity([], []); }
|
| 1220 |
+
function animateDensity() {
|
| 1221 |
+
if (densityAnim) clearInterval(densityAnim);
|
| 1222 |
+
densityAccepted = []; densityRejected = [];
|
| 1223 |
+
let count = 0;
|
| 1224 |
+
densityAnim = setInterval(() => {
|
| 1225 |
+
if (count >= 120) { clearInterval(densityAnim); return; }
|
| 1226 |
+
const p = genSyntheticDensityPt(count);
|
| 1227 |
+
if (p.accepted) densityAccepted.push(p); else densityRejected.push(p);
|
| 1228 |
+
drawDensity(densityAccepted, densityRejected);
|
| 1229 |
+
count++;
|
| 1230 |
+
}, 35);
|
| 1231 |
+
}
|
| 1232 |
+
function resetDensity() {
|
| 1233 |
+
if (densityAnim) clearInterval(densityAnim);
|
| 1234 |
+
densityAccepted = []; densityRejected = [];
|
| 1235 |
+
drawDensity([], []);
|
| 1236 |
+
}
|
| 1237 |
+
|
| 1238 |
+
// ============================================================
|
| 1239 |
+
// PANEL 5: BEFORE / AFTER
|
| 1240 |
+
// ============================================================
|
| 1241 |
+
const classes = [
|
| 1242 |
+
{ name: 'smurf', before: 2807886, after: 258783 },
|
| 1243 |
+
{ name: 'neptune', before: 1072017, after: 269180 },
|
| 1244 |
+
{ name: 'normal', before: 972781, after: 284484 },
|
| 1245 |
+
{ name: 'satan', before: 15892, after: 268453 },
|
| 1246 |
+
{ name: 'ipsweep', before: 12481, after: 270572 },
|
| 1247 |
+
{ name: 'portsweep', before: 10413, after: 268905 },
|
| 1248 |
+
{ name: 'nmap', before: 2316, after: 262351 },
|
| 1249 |
+
{ name: 'back', before: 2203, after: 256081 },
|
| 1250 |
+
{ name: 'warezclient', before: 1020, after: 264107 },
|
| 1251 |
+
{ name: 'teardrop', before: 979, after: 252848 },
|
| 1252 |
+
{ name: 'pod', before: 264, after: 240579 },
|
| 1253 |
+
];
|
| 1254 |
+
|
| 1255 |
+
function buildBars(containerId, key, palette) {
|
| 1256 |
+
const el = document.getElementById(containerId);
|
| 1257 |
+
if (!el) return;
|
| 1258 |
+
const maxVal = Math.max(...classes.map(c => c[key]));
|
| 1259 |
+
el.innerHTML = '';
|
| 1260 |
+
classes.forEach((c, i) => {
|
| 1261 |
+
const pct = (c[key] / maxVal * 100).toFixed(1);
|
| 1262 |
+
const color = key === 'before'
|
| 1263 |
+
? (c.before > 100000 ? '#ff6b35' : c.before > 10000 ? '#ffd740' : '#ff1744')
|
| 1264 |
+
: '#00e676';
|
| 1265 |
+
const row = document.createElement('div');
|
| 1266 |
+
row.className = 'bar-row';
|
| 1267 |
+
row.innerHTML = `
|
| 1268 |
+
<div class="bar-name">${c.name}</div>
|
| 1269 |
+
<div class="bar-track">
|
| 1270 |
+
<div class="bar-fill" id="bar-${key}-${i}" style="width:0%;background:${color}"></div>
|
| 1271 |
+
</div>
|
| 1272 |
+
<div class="bar-val">${(c[key] / 1000).toFixed(0)}K</div>
|
| 1273 |
+
`;
|
| 1274 |
+
el.appendChild(row);
|
| 1275 |
+
});
|
| 1276 |
+
}
|
| 1277 |
+
|
| 1278 |
+
function initBars() {
|
| 1279 |
+
buildBars('before-bars', 'before');
|
| 1280 |
+
buildBars('after-bars', 'after');
|
| 1281 |
+
}
|
| 1282 |
+
|
| 1283 |
+
function animateBars() {
|
| 1284 |
+
const maxB = Math.max(...classes.map(c => c.before));
|
| 1285 |
+
const maxA = Math.max(...classes.map(c => c.after));
|
| 1286 |
+
classes.forEach((c, i) => {
|
| 1287 |
+
setTimeout(() => {
|
| 1288 |
+
const bef = document.getElementById(`bar-before-${i}`);
|
| 1289 |
+
const aft = document.getElementById(`bar-after-${i}`);
|
| 1290 |
+
if (bef) bef.style.width = (c.before / maxB * 100) + '%';
|
| 1291 |
+
if (aft) aft.style.width = (c.after / maxA * 100) + '%';
|
| 1292 |
+
}, i * 60);
|
| 1293 |
+
});
|
| 1294 |
+
}
|
| 1295 |
+
|
| 1296 |
+
function resetBars() {
|
| 1297 |
+
classes.forEach((c, i) => {
|
| 1298 |
+
const bef = document.getElementById(`bar-before-${i}`);
|
| 1299 |
+
const aft = document.getElementById(`bar-after-${i}`);
|
| 1300 |
+
if (bef) bef.style.width = '0%';
|
| 1301 |
+
if (aft) aft.style.width = '0%';
|
| 1302 |
+
});
|
| 1303 |
+
}
|
| 1304 |
+
|
| 1305 |
+
// F1 Chart
|
| 1306 |
+
const f1Data = [
|
| 1307 |
+
{ method: 'ROS', color: '#546e7a', min: 0.002, max: 0.096 },
|
| 1308 |
+
{ method: 'SMOTE', color: '#7c4dff', min: 0.007, max: 0.096 },
|
| 1309 |
+
{ method: 'B-SMOTE', color: '#ff6b35', min: 0.016, max: 0.096 },
|
| 1310 |
+
{ method: 'ADASYN', color: '#ffd740', min: 0.017, max: 0.114 },
|
| 1311 |
+
{ method: 'DSMOTE', color: '#00e5ff', min: 0.306, max: 0.588 },
|
| 1312 |
+
];
|
| 1313 |
+
|
| 1314 |
+
let f1Anim = null, f1Progress = 0;
|
| 1315 |
+
|
| 1316 |
+
function initF1() { drawF1(0); }
|
| 1317 |
+
function animateF1() {
|
| 1318 |
+
if (f1Anim) cancelAnimationFrame(f1Anim);
|
| 1319 |
+
f1Progress = 0;
|
| 1320 |
+
function step() {
|
| 1321 |
+
f1Progress = Math.min(f1Progress + 0.025, 1);
|
| 1322 |
+
drawF1(f1Progress);
|
| 1323 |
+
if (f1Progress < 1) f1Anim = requestAnimationFrame(step);
|
| 1324 |
+
}
|
| 1325 |
+
step();
|
| 1326 |
+
}
|
| 1327 |
+
|
| 1328 |
+
function drawF1(progress) {
|
| 1329 |
+
const canvas = document.getElementById('f1-canvas');
|
| 1330 |
+
if (!canvas) return;
|
| 1331 |
+
const ctx = canvas.getContext('2d');
|
| 1332 |
+
const W = canvas.width, H = canvas.height;
|
| 1333 |
+
ctx.clearRect(0, 0, W, H);
|
| 1334 |
+
|
| 1335 |
+
const padL = 60, padR = 30, padT = 30, padB = 50;
|
| 1336 |
+
const chartW = W - padL - padR, chartH = H - padT - padB;
|
| 1337 |
+
const maxF1 = 0.65;
|
| 1338 |
+
|
| 1339 |
+
// grid
|
| 1340 |
+
ctx.strokeStyle = 'rgba(0,229,255,0.08)';
|
| 1341 |
+
ctx.lineWidth = 1;
|
| 1342 |
+
for (let i = 0; i <= 6; i++) {
|
| 1343 |
+
const y = padT + chartH * (1 - i / 6 * maxF1 / maxF1);
|
| 1344 |
+
const val = (i / 6 * maxF1).toFixed(2);
|
| 1345 |
+
ctx.beginPath(); ctx.moveTo(padL, padT + chartH - i * chartH / 6); ctx.lineTo(W - padR, padT + chartH - i * chartH / 6); ctx.stroke();
|
| 1346 |
+
ctx.fillStyle = 'rgba(84,110,122,0.7)';
|
| 1347 |
+
ctx.font = '9px JetBrains Mono, monospace';
|
| 1348 |
+
ctx.textAlign = 'right';
|
| 1349 |
+
ctx.fillText(val, padL - 6, padT + chartH - i * chartH / 6 + 3);
|
| 1350 |
+
}
|
| 1351 |
+
|
| 1352 |
+
// axes
|
| 1353 |
+
ctx.strokeStyle = 'rgba(0,229,255,0.2)';
|
| 1354 |
+
ctx.beginPath(); ctx.moveTo(padL, padT); ctx.lineTo(padL, padT + chartH); ctx.stroke();
|
| 1355 |
+
ctx.beginPath(); ctx.moveTo(padL, padT + chartH); ctx.lineTo(W - padR, padT + chartH); ctx.stroke();
|
| 1356 |
+
|
| 1357 |
+
// axis label
|
| 1358 |
+
ctx.fillStyle = 'rgba(84,110,122,0.7)';
|
| 1359 |
+
ctx.font = '10px JetBrains Mono, monospace';
|
| 1360 |
+
ctx.textAlign = 'left';
|
| 1361 |
+
ctx.fillText('Macro-F1', padL + 5, padT + 12);
|
| 1362 |
+
|
| 1363 |
+
const bw = chartW / f1Data.length;
|
| 1364 |
+
|
| 1365 |
+
f1Data.forEach((d, i) => {
|
| 1366 |
+
const x = padL + i * bw + bw * 0.15;
|
| 1367 |
+
const bWidth = bw * 0.7;
|
| 1368 |
+
|
| 1369 |
+
// range bar (min to max)
|
| 1370 |
+
const yMax = padT + chartH - (d.max * progress / maxF1) * chartH;
|
| 1371 |
+
const yMin = padT + chartH - (d.min * progress / maxF1) * chartH;
|
| 1372 |
+
|
| 1373 |
+
ctx.fillStyle = d.color + '33';
|
| 1374 |
+
ctx.fillRect(x, yMax, bWidth, yMin - yMax);
|
| 1375 |
+
ctx.strokeStyle = d.color;
|
| 1376 |
+
ctx.lineWidth = 1.5;
|
| 1377 |
+
ctx.strokeRect(x, yMax, bWidth, yMin - yMax);
|
| 1378 |
+
|
| 1379 |
+
// max line
|
| 1380 |
+
ctx.strokeStyle = d.color;
|
| 1381 |
+
ctx.lineWidth = 2.5;
|
| 1382 |
+
ctx.beginPath(); ctx.moveTo(x, yMax); ctx.lineTo(x + bWidth, yMax); ctx.stroke();
|
| 1383 |
+
|
| 1384 |
+
// value labels
|
| 1385 |
+
if (progress > 0.5) {
|
| 1386 |
+
ctx.fillStyle = d.color;
|
| 1387 |
+
ctx.font = 'bold 10px JetBrains Mono, monospace';
|
| 1388 |
+
ctx.textAlign = 'center';
|
| 1389 |
+
ctx.fillText((d.max * progress).toFixed(3), x + bWidth / 2, yMax - 5);
|
| 1390 |
+
ctx.fillStyle = 'rgba(84,110,122,0.7)';
|
| 1391 |
+
ctx.font = '9px JetBrains Mono, monospace';
|
| 1392 |
+
ctx.fillText((d.min * progress).toFixed(3), x + bWidth / 2, yMin + 12);
|
| 1393 |
+
}
|
| 1394 |
+
|
| 1395 |
+
// x label
|
| 1396 |
+
ctx.fillStyle = i === 4 ? d.color : 'rgba(84,110,122,0.8)';
|
| 1397 |
+
ctx.font = i === 4 ? 'bold 10px JetBrains Mono, monospace' : '9px JetBrains Mono, monospace';
|
| 1398 |
+
ctx.textAlign = 'center';
|
| 1399 |
+
ctx.fillText(d.method, x + bWidth / 2, padT + chartH + 20);
|
| 1400 |
+
if (i === 4) ctx.fillText('★', x + bWidth / 2, padT + chartH + 33);
|
| 1401 |
+
});
|
| 1402 |
+
|
| 1403 |
+
// DSMOTE highlight
|
| 1404 |
+
if (progress > 0.7) {
|
| 1405 |
+
ctx.strokeStyle = 'rgba(0,229,255,0.3)';
|
| 1406 |
+
ctx.lineWidth = 1;
|
| 1407 |
+
ctx.setLineDash([4, 4]);
|
| 1408 |
+
const dsmX = padL + 4 * bw;
|
| 1409 |
+
ctx.strokeRect(dsmX, padT, bw, chartH);
|
| 1410 |
+
ctx.setLineDash([]);
|
| 1411 |
+
}
|
| 1412 |
+
ctx.textAlign = 'left';
|
| 1413 |
+
}
|
| 1414 |
+
|
| 1415 |
+
// ============================================================
|
| 1416 |
+
// PANEL 6: UNSW RESULTS
|
| 1417 |
+
// ============================================================
|
| 1418 |
+
|
| 1419 |
+
const unswModels = ['DT', 'RF', 'XGBoost', 'ANN', 'CNN', 'LSTM', 'LSTM-CNN'];
|
| 1420 |
+
const unswMethods = [
|
| 1421 |
+
{ name: 'RAW', color: '#78909c', f1: [0.485, 0.581, 0.451, 0.299, 0.273, 0.325, 0.395] },
|
| 1422 |
+
{ name: 'ROS', color: '#546e7a', f1: [0.096, 0.096, 0.003, 0.096, 0.096, 0.026, 0.016] },
|
| 1423 |
+
{ name: 'SMOTE', color: '#7c4dff', f1: [0.094, 0.096, 0.025, 0.096, 0.096, 0.010, 0.008] },
|
| 1424 |
+
{ name: 'BSMOTE', color: '#ff6b35', f1: [0.061, 0.096, 0.040, 0.096, 0.096, 0.016, 0.022] },
|
| 1425 |
+
{ name: 'ADASYN', color: '#ffd740', f1: [0.062, 0.097, 0.114, 0.096, 0.096, 0.017, 0.018] },
|
| 1426 |
+
{ name: 'DSMOTE', color: '#00e5ff', f1: [0.535, 0.588, 0.432, 0.307, 0.274, 0.332, 0.448] },
|
| 1427 |
+
];
|
| 1428 |
+
|
| 1429 |
+
let unswFilter = 'ALL';
|
| 1430 |
+
let unswAnimProgress = 0, unswAnimFrame = null;
|
| 1431 |
+
|
| 1432 |
+
function filterUnswModel(model) {
|
| 1433 |
+
unswFilter = model;
|
| 1434 |
+
unswAnimProgress = 0;
|
| 1435 |
+
if (unswAnimFrame) cancelAnimationFrame(unswAnimFrame);
|
| 1436 |
+
function step() {
|
| 1437 |
+
unswAnimProgress = Math.min(unswAnimProgress + 0.04, 1);
|
| 1438 |
+
drawUnswF1(unswAnimProgress);
|
| 1439 |
+
if (unswAnimProgress < 1) unswAnimFrame = requestAnimationFrame(step);
|
| 1440 |
+
}
|
| 1441 |
+
step();
|
| 1442 |
+
}
|
| 1443 |
+
|
| 1444 |
+
function drawUnswF1(progress) {
|
| 1445 |
+
const canvas = document.getElementById('unsw-f1-canvas');
|
| 1446 |
+
if (!canvas) return;
|
| 1447 |
+
const ctx = canvas.getContext('2d');
|
| 1448 |
+
const W = canvas.width, H = canvas.height;
|
| 1449 |
+
ctx.clearRect(0, 0, W, H);
|
| 1450 |
+
|
| 1451 |
+
const models = unswFilter === 'ALL' ? unswModels : [unswFilter];
|
| 1452 |
+
const modelIndices = unswFilter === 'ALL'
|
| 1453 |
+
? unswModels.map((_, i) => i)
|
| 1454 |
+
: [unswModels.indexOf(unswFilter)];
|
| 1455 |
+
|
| 1456 |
+
const padL = 45, padR = 20, padT = 30, padB = 55;
|
| 1457 |
+
const chartW = W - padL - padR, chartH = H - padT - padB;
|
| 1458 |
+
const maxF1 = 0.65;
|
| 1459 |
+
|
| 1460 |
+
// Grid
|
| 1461 |
+
ctx.strokeStyle = 'rgba(0,229,255,0.08)'; ctx.lineWidth = 1;
|
| 1462 |
+
for (let i = 0; i <= 6; i++) {
|
| 1463 |
+
const y = padT + chartH * (1 - i / 6);
|
| 1464 |
+
ctx.beginPath(); ctx.moveTo(padL, y); ctx.lineTo(W - padR, y); ctx.stroke();
|
| 1465 |
+
ctx.fillStyle = 'rgba(84,110,122,0.7)';
|
| 1466 |
+
ctx.font = '9px JetBrains Mono, monospace';
|
| 1467 |
+
ctx.textAlign = 'right';
|
| 1468 |
+
ctx.fillText((maxF1 * i / 6).toFixed(2), padL - 5, y + 3);
|
| 1469 |
+
}
|
| 1470 |
+
ctx.strokeStyle = 'rgba(0,229,255,0.2)';
|
| 1471 |
+
ctx.beginPath(); ctx.moveTo(padL, padT); ctx.lineTo(padL, padT + chartH); ctx.stroke();
|
| 1472 |
+
ctx.beginPath(); ctx.moveTo(padL, padT + chartH); ctx.lineTo(W - padR, padT + chartH); ctx.stroke();
|
| 1473 |
+
|
| 1474 |
+
const groupW = chartW / models.length;
|
| 1475 |
+
const nMethods = unswMethods.length;
|
| 1476 |
+
const bw = groupW * 0.8 / nMethods;
|
| 1477 |
+
const groupPad = groupW * 0.1;
|
| 1478 |
+
|
| 1479 |
+
modelIndices.forEach((mi, gi) => {
|
| 1480 |
+
const gx = padL + gi * groupW + groupPad;
|
| 1481 |
+
// model label
|
| 1482 |
+
ctx.fillStyle = 'rgba(224,247,250,0.7)';
|
| 1483 |
+
ctx.font = 'bold 10px JetBrains Mono, monospace';
|
| 1484 |
+
ctx.textAlign = 'center';
|
| 1485 |
+
ctx.fillText(unswModels[mi], padL + gi * groupW + groupW / 2, padT + chartH + 20);
|
| 1486 |
+
|
| 1487 |
+
unswMethods.forEach((m, mIdx) => {
|
| 1488 |
+
const val = m.f1[mi] * progress;
|
| 1489 |
+
const barH = (val / maxF1) * chartH;
|
| 1490 |
+
const x = gx + mIdx * bw;
|
| 1491 |
+
const y = padT + chartH - barH;
|
| 1492 |
+
const isDSMOTE = m.name === 'DSMOTE';
|
| 1493 |
+
|
| 1494 |
+
ctx.fillStyle = isDSMOTE ? m.color + 'cc' : m.color + '88';
|
| 1495 |
+
ctx.fillRect(x, y, bw - 1, barH);
|
| 1496 |
+
if (isDSMOTE) {
|
| 1497 |
+
ctx.strokeStyle = m.color;
|
| 1498 |
+
ctx.lineWidth = 1.5;
|
| 1499 |
+
ctx.strokeRect(x, y, bw - 1, barH);
|
| 1500 |
+
}
|
| 1501 |
+
|
| 1502 |
+
if (progress > 0.85 && val > 0.05) {
|
| 1503 |
+
ctx.fillStyle = isDSMOTE ? m.color : 'rgba(84,110,122,0.8)';
|
| 1504 |
+
ctx.font = isDSMOTE ? 'bold 8px JetBrains Mono,monospace' : '8px JetBrains Mono,monospace';
|
| 1505 |
+
ctx.textAlign = 'center';
|
| 1506 |
+
ctx.fillText(val.toFixed(2), x + (bw - 1) / 2, y - 3);
|
| 1507 |
+
}
|
| 1508 |
+
});
|
| 1509 |
+
});
|
| 1510 |
+
|
| 1511 |
+
// Method legend inside chart
|
| 1512 |
+
ctx.textAlign = 'left';
|
| 1513 |
+
}
|
| 1514 |
+
|
| 1515 |
+
// UNSW Balanced Accuracy
|
| 1516 |
+
const unswBA = [
|
| 1517 |
+
{ method: 'RAW', color: '#78909c', vals: [0.470, 0.566, 0.425, 0.374, 0.278, 0.329, 0.382] },
|
| 1518 |
+
{ method: 'SMOTE', color: '#7c4dff', vals: [0.096, 0.100, 0.111, 0.100, 0.100, 0.068, 0.124] },
|
| 1519 |
+
{ method: 'ADASYN', color: '#ffd740', vals: [0.050, 0.101, 0.168, 0.100, 0.100, 0.082, 0.133] },
|
| 1520 |
+
{ method: 'DSMOTE', color: '#00e5ff', vals: [0.516, 0.573, 0.408, 0.376, 0.290, 0.332, 0.479] },
|
| 1521 |
+
];
|
| 1522 |
+
|
| 1523 |
+
function drawUnswBA() {
|
| 1524 |
+
const canvas = document.getElementById('unsw-ba-canvas');
|
| 1525 |
+
if (!canvas) return;
|
| 1526 |
+
const ctx = canvas.getContext('2d');
|
| 1527 |
+
const W = canvas.width, H = canvas.height;
|
| 1528 |
+
ctx.clearRect(0, 0, W, H);
|
| 1529 |
+
const padL = 45, padR = 20, padT = 25, padB = 50;
|
| 1530 |
+
const chartW = W - padL - padR, chartH = H - padT - padB;
|
| 1531 |
+
|
| 1532 |
+
ctx.strokeStyle = 'rgba(0,229,255,0.08)'; ctx.lineWidth = 1;
|
| 1533 |
+
for (let i = 0; i <= 5; i++) {
|
| 1534 |
+
const y = padT + chartH * (1 - i / 5);
|
| 1535 |
+
ctx.beginPath(); ctx.moveTo(padL, y); ctx.lineTo(W - padR, y); ctx.stroke();
|
| 1536 |
+
ctx.fillStyle = 'rgba(84,110,122,0.7)'; ctx.font = '9px JetBrains Mono,monospace'; ctx.textAlign = 'right';
|
| 1537 |
+
ctx.fillText((i / 5).toFixed(1), padL - 4, y + 3);
|
| 1538 |
+
}
|
| 1539 |
+
|
| 1540 |
+
const groupW = chartW / unswModels.length;
|
| 1541 |
+
unswModels.forEach((model, mi) => {
|
| 1542 |
+
const gx = padL + mi * groupW + groupW * 0.08;
|
| 1543 |
+
const bw = groupW * 0.84 / unswBA.length;
|
| 1544 |
+
ctx.fillStyle = 'rgba(224,247,250,0.7)'; ctx.font = '10px JetBrains Mono,monospace';
|
| 1545 |
+
ctx.textAlign = 'center'; ctx.fillText(model, padL + mi * groupW + groupW / 2, padT + chartH + 18);
|
| 1546 |
+
unswBA.forEach((m, mIdx) => {
|
| 1547 |
+
const val = m.vals[mi];
|
| 1548 |
+
const barH = val * chartH;
|
| 1549 |
+
const x = gx + mIdx * bw;
|
| 1550 |
+
const y = padT + chartH - barH;
|
| 1551 |
+
const isDSMOTE = m.method === 'DSMOTE';
|
| 1552 |
+
ctx.fillStyle = isDSMOTE ? m.color + 'cc' : m.color + '77';
|
| 1553 |
+
ctx.fillRect(x, y, bw - 1, barH);
|
| 1554 |
+
if (isDSMOTE) {
|
| 1555 |
+
ctx.strokeStyle = m.color; ctx.lineWidth = 1.5;
|
| 1556 |
+
ctx.strokeRect(x, y, bw - 1, barH);
|
| 1557 |
+
ctx.fillStyle = m.color; ctx.font = 'bold 8px JetBrains Mono,monospace';
|
| 1558 |
+
ctx.fillText(val.toFixed(2), x + (bw - 1) / 2, y - 3);
|
| 1559 |
+
}
|
| 1560 |
+
});
|
| 1561 |
+
});
|
| 1562 |
+
ctx.textAlign = 'left';
|
| 1563 |
+
}
|
| 1564 |
+
|
| 1565 |
+
// Confusion Matrix renderer
|
| 1566 |
+
function drawConfusionMatrices() {
|
| 1567 |
+
drawCM('cm-smote-canvas', 'SMOTE RF', false);
|
| 1568 |
+
drawCM('cm-dsmote-canvas', 'DSMOTE RF', true);
|
| 1569 |
+
}
|
| 1570 |
+
|
| 1571 |
+
function drawCM(canvasId, title, isDsmote) {
|
| 1572 |
+
const canvas = document.getElementById(canvasId);
|
| 1573 |
+
if (!canvas) return;
|
| 1574 |
+
const ctx = canvas.getContext('2d');
|
| 1575 |
+
const W = canvas.width, H = canvas.height;
|
| 1576 |
+
ctx.clearRect(0, 0, W, H);
|
| 1577 |
+
|
| 1578 |
+
const classes = ['Benign','Exploits','Fuzzers','Recon.','Generic','DoS','Backdoor','Shellcode','Analysis','Worms'];
|
| 1579 |
+
const n = classes.length;
|
| 1580 |
+
|
| 1581 |
+
// SMOTE RF: predicts everything as Benign (col 0)
|
| 1582 |
+
// DSMOTE RF: approximate from RAW_RF confusion data (spread across diagonal)
|
| 1583 |
+
let matrix;
|
| 1584 |
+
if (!isDsmote) {
|
| 1585 |
+
// SMOTE collapse: everything predicted as Benign
|
| 1586 |
+
matrix = Array.from({length:n}, (_, r) => Array.from({length:n}, (_, c) => c === 0 ? [184,583,169987,805,6073,4048,1139,1787,253,21][r] : 0));
|
| 1587 |
+
} else {
|
| 1588 |
+
// DSMOTE / RAW RF-like: diagonal dominant with some spread
|
| 1589 |
+
matrix = [
|
| 1590 |
+
[67,0,0,4,105,6,0,2,0,0],
|
| 1591 |
+
[0,57,0,16,202,201,60,35,10,2],
|
| 1592 |
+
[0,1,169977,3,5,1,0,0,0,0],
|
| 1593 |
+
[11,3,1,246,324,165,25,23,8,0],
|
| 1594 |
+
[69,25,0,89,4780,752,86,229,39,4],
|
| 1595 |
+
[36,47,0,31,304,3485,50,91,4,0],
|
| 1596 |
+
[6,43,0,15,148,133,731,51,13,1],
|
| 1597 |
+
[23,13,0,18,342,404,43,921,23,0],
|
| 1598 |
+
[0,9,0,2,53,45,11,27,106,0],
|
| 1599 |
+
[0,0,0,1,5,0,0,0,1,14],
|
| 1600 |
+
];
|
| 1601 |
+
}
|
| 1602 |
+
|
| 1603 |
+
const padL = 70, padT = 30, padR = 15, padB = 70;
|
| 1604 |
+
const cellW = (W - padL - padR) / n;
|
| 1605 |
+
const cellH = (H - padT - padB) / n;
|
| 1606 |
+
|
| 1607 |
+
// Normalize
|
| 1608 |
+
const rowMaxes = matrix.map(row => Math.max(...row));
|
| 1609 |
+
const globalMax = Math.max(...rowMaxes);
|
| 1610 |
+
|
| 1611 |
+
matrix.forEach((row, ri) => {
|
| 1612 |
+
row.forEach((val, ci) => {
|
| 1613 |
+
const x = padL + ci * cellW;
|
| 1614 |
+
const y = padT + ri * cellH;
|
| 1615 |
+
const intensity = val / globalMax;
|
| 1616 |
+
const color = isDsmote
|
| 1617 |
+
? `rgba(0,229,255,${Math.min(intensity * 1.5, 0.9)})`
|
| 1618 |
+
: `rgba(255,23,68,${Math.min(intensity * 1.5, 0.9)})`;
|
| 1619 |
+
ctx.fillStyle = intensity > 0.01 ? color : 'rgba(6,11,20,0.8)';
|
| 1620 |
+
ctx.fillRect(x + 1, y + 1, cellW - 2, cellH - 2);
|
| 1621 |
+
// value
|
| 1622 |
+
if (val > 0) {
|
| 1623 |
+
ctx.fillStyle = intensity > 0.3 ? 'rgba(6,11,20,0.9)' : 'rgba(224,247,250,0.7)';
|
| 1624 |
+
ctx.font = `${Math.min(cellW * 0.28, 10)}px JetBrains Mono,monospace`;
|
| 1625 |
+
ctx.textAlign = 'center';
|
| 1626 |
+
const display = val > 1000 ? (val/1000).toFixed(0)+'k' : val;
|
| 1627 |
+
ctx.fillText(display, x + cellW/2, y + cellH/2 + 3);
|
| 1628 |
+
}
|
| 1629 |
+
});
|
| 1630 |
+
});
|
| 1631 |
+
|
| 1632 |
+
// X labels
|
| 1633 |
+
ctx.fillStyle = 'rgba(84,110,122,0.9)'; ctx.font = '8px JetBrains Mono,monospace'; ctx.textAlign = 'center';
|
| 1634 |
+
classes.forEach((c, i) => {
|
| 1635 |
+
ctx.save(); ctx.translate(padL + i * cellW + cellW/2, H - padB + 8);
|
| 1636 |
+
ctx.rotate(-Math.PI / 4); ctx.fillText(c, 0, 0); ctx.restore();
|
| 1637 |
+
});
|
| 1638 |
+
// Y labels
|
| 1639 |
+
ctx.textAlign = 'right';
|
| 1640 |
+
classes.forEach((c, i) => {
|
| 1641 |
+
ctx.fillText(c, padL - 4, padT + i * cellH + cellH/2 + 3);
|
| 1642 |
+
});
|
| 1643 |
+
|
| 1644 |
+
// Title
|
| 1645 |
+
ctx.fillStyle = isDsmote ? '#00e5ff' : '#ff1744';
|
| 1646 |
+
ctx.font = 'bold 11px JetBrains Mono,monospace'; ctx.textAlign = 'center';
|
| 1647 |
+
ctx.fillText(title, W/2, 18);
|
| 1648 |
+
ctx.textAlign = 'left';
|
| 1649 |
+
}
|
| 1650 |
+
|
| 1651 |
+
// ============================================================
|
| 1652 |
+
// PANEL 7: KDD RESULTS
|
| 1653 |
+
// ============================================================
|
| 1654 |
+
const kddModels = ['DT', 'RF', 'XGBoost', 'ANN', 'CNN', 'LSTM', 'LSTM-CNN'];
|
| 1655 |
+
const kddMethods = [
|
| 1656 |
+
{ name: 'RAW', color: '#78909c', f1: [0.962, 0.995, 0.992, 0.952, 0.645, 0.952, 0.976] },
|
| 1657 |
+
{ name: 'ROS', color: '#546e7a', f1: [0.925, 0.996, 0.992, 0.858, 0.627, 0.858, 0.975] },
|
| 1658 |
+
{ name: 'SMOTE', color: '#7c4dff', f1: [0.939, 0.995, 0.992, 0.866, 0.620, 0.943, 0.974] },
|
| 1659 |
+
{ name: 'BSMOTE', color: '#ff6b35', f1: [0.911, 0.995, 0.991, 0.900, 0.452, 0.924, 0.945] },
|
| 1660 |
+
{ name: 'DSMOTE', color: '#00e5ff', f1: [0.962, 0.995, 0.992, 0.928, 0.505, 0.944, 0.965] },
|
| 1661 |
+
];
|
| 1662 |
+
|
| 1663 |
+
let kddAnimFrame = null, kddProgress = 0;
|
| 1664 |
+
|
| 1665 |
+
function drawKddF1(progress) {
|
| 1666 |
+
const canvas = document.getElementById('kdd-f1-canvas');
|
| 1667 |
+
if (!canvas) return;
|
| 1668 |
+
const ctx = canvas.getContext('2d');
|
| 1669 |
+
const W = canvas.width, H = canvas.height;
|
| 1670 |
+
ctx.clearRect(0, 0, W, H);
|
| 1671 |
+
const padL = 45, padR = 20, padT = 30, padB = 55;
|
| 1672 |
+
const chartW = W - padL - padR, chartH = H - padT - padB;
|
| 1673 |
+
const minF1 = 0.4, maxF1 = 1.0, range = maxF1 - minF1;
|
| 1674 |
+
|
| 1675 |
+
ctx.strokeStyle = 'rgba(0,229,255,0.08)'; ctx.lineWidth = 1;
|
| 1676 |
+
for (let i = 0; i <= 6; i++) {
|
| 1677 |
+
const val = minF1 + range * i / 6;
|
| 1678 |
+
const y = padT + chartH * (1 - i / 6);
|
| 1679 |
+
ctx.beginPath(); ctx.moveTo(padL, y); ctx.lineTo(W - padR, y); ctx.stroke();
|
| 1680 |
+
ctx.fillStyle = 'rgba(84,110,122,0.7)'; ctx.font = '9px JetBrains Mono,monospace';
|
| 1681 |
+
ctx.textAlign = 'right'; ctx.fillText(val.toFixed(2), padL - 4, y + 3);
|
| 1682 |
+
}
|
| 1683 |
+
// note: zoom view
|
| 1684 |
+
ctx.fillStyle = 'rgba(84,110,122,0.5)'; ctx.font = '9px JetBrains Mono,monospace';
|
| 1685 |
+
ctx.textAlign = 'left'; ctx.fillText('* Y-axis starts at 0.40 for readability', padL + 5, padT + 12);
|
| 1686 |
+
|
| 1687 |
+
const groupW = chartW / kddModels.length;
|
| 1688 |
+
kddModels.forEach((model, mi) => {
|
| 1689 |
+
const gx = padL + mi * groupW + groupW * 0.05;
|
| 1690 |
+
const bw = groupW * 0.9 / kddMethods.length;
|
| 1691 |
+
ctx.fillStyle = 'rgba(224,247,250,0.7)'; ctx.font = '10px JetBrains Mono,monospace';
|
| 1692 |
+
ctx.textAlign = 'center'; ctx.fillText(model, padL + mi * groupW + groupW / 2, padT + chartH + 18);
|
| 1693 |
+
kddMethods.forEach((m, mIdx) => {
|
| 1694 |
+
const raw = m.f1[mi];
|
| 1695 |
+
const val = Math.max(raw - minF1, 0) * progress;
|
| 1696 |
+
const barH = (val / range) * chartH;
|
| 1697 |
+
const x = gx + mIdx * bw;
|
| 1698 |
+
const y = padT + chartH - barH;
|
| 1699 |
+
const isDSMOTE = m.name === 'DSMOTE';
|
| 1700 |
+
ctx.fillStyle = isDSMOTE ? m.color + 'cc' : m.color + '77';
|
| 1701 |
+
ctx.fillRect(x, y, bw - 1, barH);
|
| 1702 |
+
if (isDSMOTE) {
|
| 1703 |
+
ctx.strokeStyle = m.color; ctx.lineWidth = 1.5;
|
| 1704 |
+
ctx.strokeRect(x, y, bw - 1, barH);
|
| 1705 |
+
}
|
| 1706 |
+
if (progress > 0.85 && raw > minF1 + 0.05) {
|
| 1707 |
+
ctx.fillStyle = isDSMOTE ? m.color : 'rgba(84,110,122,0.7)';
|
| 1708 |
+
ctx.font = isDSMOTE ? 'bold 8px JetBrains Mono,monospace' : '8px JetBrains Mono,monospace';
|
| 1709 |
+
ctx.textAlign = 'center';
|
| 1710 |
+
ctx.fillText(raw.toFixed(3), x + (bw-1)/2, y - 3);
|
| 1711 |
+
}
|
| 1712 |
+
});
|
| 1713 |
+
});
|
| 1714 |
+
ctx.textAlign = 'left';
|
| 1715 |
+
}
|
| 1716 |
+
|
| 1717 |
+
// Radar chart
|
| 1718 |
+
function drawRadar() {
|
| 1719 |
+
const canvas = document.getElementById('kdd-radar-canvas');
|
| 1720 |
+
if (!canvas) return;
|
| 1721 |
+
const ctx = canvas.getContext('2d');
|
| 1722 |
+
const W = canvas.width, H = canvas.height;
|
| 1723 |
+
ctx.clearRect(0, 0, W, H);
|
| 1724 |
+
const cx = W / 2, cy = H / 2, R = Math.min(W, H) * 0.36;
|
| 1725 |
+
const axes = ['Accuracy', 'Bal. Acc.', 'Precision', 'Recall', 'F1 Macro', 'G-Mean'];
|
| 1726 |
+
const n = axes.length;
|
| 1727 |
+
const datasets = [
|
| 1728 |
+
{ name: 'RAW RF', color: '#78909c', vals: [0.99986, 0.99402, 0.99693, 0.99402, 0.99545, 0.99393] },
|
| 1729 |
+
{ name: 'DSMOTE RF', color: '#00e5ff', vals: [0.99986, 0.99402, 0.99693, 0.99402, 0.99545, 0.99393] },
|
| 1730 |
+
{ name: 'SMOTE RF', color: '#7c4dff', vals: [0.9999, 0.9922, 0.9978, 0.9922, 0.9949, 0.9920] },
|
| 1731 |
+
];
|
| 1732 |
+
|
| 1733 |
+
// Grid rings
|
| 1734 |
+
for (let ring = 1; ring <= 5; ring++) {
|
| 1735 |
+
const r = R * ring / 5;
|
| 1736 |
+
ctx.beginPath();
|
| 1737 |
+
for (let i = 0; i < n; i++) {
|
| 1738 |
+
const angle = (i / n) * Math.PI * 2 - Math.PI / 2;
|
| 1739 |
+
const x = cx + r * Math.cos(angle), y = cy + r * Math.sin(angle);
|
| 1740 |
+
i === 0 ? ctx.moveTo(x, y) : ctx.lineTo(x, y);
|
| 1741 |
+
}
|
| 1742 |
+
ctx.closePath();
|
| 1743 |
+
ctx.strokeStyle = 'rgba(0,229,255,0.1)'; ctx.lineWidth = 1; ctx.stroke();
|
| 1744 |
+
ctx.fillStyle = 'rgba(84,110,122,0.5)'; ctx.font = '8px JetBrains Mono,monospace'; ctx.textAlign = 'center';
|
| 1745 |
+
ctx.fillText((ring / 5).toFixed(1), cx, cy - r - 3);
|
| 1746 |
+
}
|
| 1747 |
+
|
| 1748 |
+
// Axis lines & labels
|
| 1749 |
+
for (let i = 0; i < n; i++) {
|
| 1750 |
+
const angle = (i / n) * Math.PI * 2 - Math.PI / 2;
|
| 1751 |
+
const ex = cx + R * Math.cos(angle), ey = cy + R * Math.sin(angle);
|
| 1752 |
+
ctx.beginPath(); ctx.moveTo(cx, cy); ctx.lineTo(ex, ey);
|
| 1753 |
+
ctx.strokeStyle = 'rgba(0,229,255,0.2)'; ctx.stroke();
|
| 1754 |
+
const lx = cx + (R + 22) * Math.cos(angle), ly = cy + (R + 22) * Math.sin(angle);
|
| 1755 |
+
ctx.fillStyle = 'rgba(224,247,250,0.7)'; ctx.font = '10px JetBrains Mono,monospace';
|
| 1756 |
+
ctx.textAlign = 'center'; ctx.fillText(axes[i], lx, ly + 3);
|
| 1757 |
+
}
|
| 1758 |
+
|
| 1759 |
+
// Data polygons
|
| 1760 |
+
datasets.forEach((ds, di) => {
|
| 1761 |
+
ctx.beginPath();
|
| 1762 |
+
ds.vals.forEach((v, i) => {
|
| 1763 |
+
const angle = (i / n) * Math.PI * 2 - Math.PI / 2;
|
| 1764 |
+
const r = v * R;
|
| 1765 |
+
const x = cx + r * Math.cos(angle), y = cy + r * Math.sin(angle);
|
| 1766 |
+
i === 0 ? ctx.moveTo(x, y) : ctx.lineTo(x, y);
|
| 1767 |
+
});
|
| 1768 |
+
ctx.closePath();
|
| 1769 |
+
ctx.strokeStyle = ds.color; ctx.lineWidth = di === 1 ? 2.5 : 1.5; ctx.stroke();
|
| 1770 |
+
ctx.fillStyle = ds.color + (di === 1 ? '22' : '11'); ctx.fill();
|
| 1771 |
+
// dots
|
| 1772 |
+
ds.vals.forEach((v, i) => {
|
| 1773 |
+
const angle = (i / n) * Math.PI * 2 - Math.PI / 2;
|
| 1774 |
+
const r = v * R;
|
| 1775 |
+
const x = cx + r * Math.cos(angle), y = cy + r * Math.sin(angle);
|
| 1776 |
+
ctx.beginPath(); ctx.arc(x, y, di === 1 ? 4 : 3, 0, Math.PI * 2);
|
| 1777 |
+
ctx.fillStyle = ds.color; ctx.fill();
|
| 1778 |
+
});
|
| 1779 |
+
});
|
| 1780 |
+
|
| 1781 |
+
// Legend
|
| 1782 |
+
datasets.forEach((ds, i) => {
|
| 1783 |
+
const lx = 20, ly = 20 + i * 18;
|
| 1784 |
+
ctx.fillStyle = ds.color; ctx.fillRect(lx, ly, 14, 3);
|
| 1785 |
+
ctx.fillStyle = 'rgba(224,247,250,0.8)'; ctx.font = '10px JetBrains Mono,monospace';
|
| 1786 |
+
ctx.textAlign = 'left'; ctx.fillText(ds.name, lx + 20, ly + 4);
|
| 1787 |
+
});
|
| 1788 |
+
ctx.textAlign = 'left';
|
| 1789 |
+
}
|
| 1790 |
+
|
| 1791 |
+
// KDD G-Mean
|
| 1792 |
+
const kddGmean = [
|
| 1793 |
+
{ method: 'RAW', color: '#78909c', vals: [0.959, 0.994, 0.991, 0.936, 0.005, 0.966, 0.985] },
|
| 1794 |
+
{ method: 'SMOTE', color: '#7c4dff', vals: [0.993, 0.992, 0.997, 0.991, 0.972, 0.991, 0.991] },
|
| 1795 |
+
{ method: 'BSMOTE', color: '#ff6b35', vals: [0.991, 0.992, 0.997, 0.958, 0.655, 0.958, 0.977] },
|
| 1796 |
+
{ method: 'DSMOTE', color: '#00e5ff', vals: [0.959, 0.994, 0.991, 0.943, 0.560, 0.940, 0.959] },
|
| 1797 |
+
];
|
| 1798 |
+
|
| 1799 |
+
function drawKddGmean() {
|
| 1800 |
+
const canvas = document.getElementById('kdd-gmean-canvas');
|
| 1801 |
+
if (!canvas) return;
|
| 1802 |
+
const ctx = canvas.getContext('2d');
|
| 1803 |
+
const W = canvas.width, H = canvas.height;
|
| 1804 |
+
ctx.clearRect(0, 0, W, H);
|
| 1805 |
+
const padL = 45, padR = 20, padT = 25, padB = 50;
|
| 1806 |
+
const chartW = W - padL - padR, chartH = H - padT - padB;
|
| 1807 |
+
const minV = 0.0, maxV = 1.0;
|
| 1808 |
+
|
| 1809 |
+
ctx.strokeStyle = 'rgba(0,229,255,0.08)'; ctx.lineWidth = 1;
|
| 1810 |
+
for (let i = 0; i <= 5; i++) {
|
| 1811 |
+
const y = padT + chartH * (1 - i / 5);
|
| 1812 |
+
ctx.beginPath(); ctx.moveTo(padL, y); ctx.lineTo(W - padR, y); ctx.stroke();
|
| 1813 |
+
ctx.fillStyle = 'rgba(84,110,122,0.7)'; ctx.font = '9px JetBrains Mono,monospace';
|
| 1814 |
+
ctx.textAlign = 'right'; ctx.fillText((i / 5).toFixed(1), padL - 4, y + 3);
|
| 1815 |
+
}
|
| 1816 |
+
|
| 1817 |
+
const groupW = chartW / kddModels.length;
|
| 1818 |
+
kddModels.forEach((model, mi) => {
|
| 1819 |
+
const gx = padL + mi * groupW + groupW * 0.06;
|
| 1820 |
+
const bw = groupW * 0.88 / kddGmean.length;
|
| 1821 |
+
ctx.fillStyle = 'rgba(224,247,250,0.7)'; ctx.font = '10px JetBrains Mono,monospace';
|
| 1822 |
+
ctx.textAlign = 'center'; ctx.fillText(model, padL + mi * groupW + groupW / 2, padT + chartH + 18);
|
| 1823 |
+
kddGmean.forEach((m, mIdx) => {
|
| 1824 |
+
const val = m.vals[mi];
|
| 1825 |
+
const barH = val * chartH;
|
| 1826 |
+
const x = gx + mIdx * bw;
|
| 1827 |
+
const y = padT + chartH - barH;
|
| 1828 |
+
const isDSMOTE = m.method === 'DSMOTE';
|
| 1829 |
+
ctx.fillStyle = isDSMOTE ? m.color + 'cc' : m.color + '77';
|
| 1830 |
+
ctx.fillRect(x, y, bw - 1, barH);
|
| 1831 |
+
if (isDSMOTE) {
|
| 1832 |
+
ctx.strokeStyle = m.color; ctx.lineWidth = 1.5;
|
| 1833 |
+
ctx.strokeRect(x, y, bw - 1, barH);
|
| 1834 |
+
ctx.fillStyle = m.color; ctx.font = 'bold 8px JetBrains Mono,monospace';
|
| 1835 |
+
ctx.textAlign = 'center'; ctx.fillText(val.toFixed(3), x + (bw-1)/2, y - 3);
|
| 1836 |
+
}
|
| 1837 |
+
});
|
| 1838 |
+
});
|
| 1839 |
+
ctx.textAlign = 'left';
|
| 1840 |
+
}
|
| 1841 |
+
|
| 1842 |
+
// ============================================================
|
| 1843 |
+
// INIT
|
| 1844 |
+
// ============================================================
|
| 1845 |
+
window.addEventListener('load', () => {
|
| 1846 |
+
resetSmote();
|
| 1847 |
+
drawScatterBase('smote-canvas', [], '#ff6b35');
|
| 1848 |
+
drawScatterBase('dsmote-canvas', [], '#00e676');
|
| 1849 |
+
});
|
| 1850 |
+
|
| 1851 |
+
// Auto-draw on tab switch
|
| 1852 |
+
const _origSwitchTab = switchTab;
|
| 1853 |
+
function switchTab(i) {
|
| 1854 |
+
document.querySelectorAll('.tab').forEach((t, j) => t.classList.toggle('active', i === j));
|
| 1855 |
+
document.querySelectorAll('.panel').forEach((p, j) => p.classList.toggle('active', i === j));
|
| 1856 |
+
if (i === 1) setTimeout(initPipeline, 100);
|
| 1857 |
+
if (i === 2) setTimeout(initClusters, 100);
|
| 1858 |
+
if (i === 3) setTimeout(initDensity, 100);
|
| 1859 |
+
if (i === 4) setTimeout(() => { initBars(); initF1(); }, 100);
|
| 1860 |
+
if (i === 5) {
|
| 1861 |
+
setTimeout(() => {
|
| 1862 |
+
filterUnswModel('ALL');
|
| 1863 |
+
drawUnswBA();
|
| 1864 |
+
drawConfusionMatrices();
|
| 1865 |
+
}, 150);
|
| 1866 |
+
}
|
| 1867 |
+
if (i === 6) {
|
| 1868 |
+
setTimeout(() => {
|
| 1869 |
+
kddProgress = 0;
|
| 1870 |
+
if (kddAnimFrame) cancelAnimationFrame(kddAnimFrame);
|
| 1871 |
+
function step() {
|
| 1872 |
+
kddProgress = Math.min(kddProgress + 0.04, 1);
|
| 1873 |
+
drawKddF1(kddProgress);
|
| 1874 |
+
if (kddProgress < 1) kddAnimFrame = requestAnimationFrame(step);
|
| 1875 |
+
}
|
| 1876 |
+
step();
|
| 1877 |
+
drawRadar();
|
| 1878 |
+
drawKddGmean();
|
| 1879 |
+
}, 150);
|
| 1880 |
+
}
|
| 1881 |
+
}
|
| 1882 |
+
</script>
|
| 1883 |
+
</body>
|
| 1884 |
</html>
|