Spaces:

eyad222
/

NIDS_Survey

Running

File size: 41,832 Bytes

<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>From Data to Defense — NIDS Survey Results</title>
<link href="https://fonts.googleapis.com/css2?family=Share+Tech+Mono&family=Sora:wght@300;400;600;700;800&display=swap" rel="stylesheet">
<style>
  :root {
    --bg: #040c14;
    --bg2: #071622;
    --bg3: #0a1f2e;
    --cyan: #00d4ff;
    --cyan2: #00a8cc;
    --green: #00ff9d;
    --orange: #ff6b35;
    --yellow: #ffd700;
    --red: #ff3d5a;
    --text: #d4eaf7;
    --muted: #5a7a8a;
    --border: rgba(0,212,255,0.15);
    --mono: 'Share Tech Mono', monospace;
    --sans: 'Sora', sans-serif;
  }

  * { margin: 0; padding: 0; box-sizing: border-box; }

  html { scroll-behavior: smooth; }

  body {
    background: var(--bg);
    color: var(--text);
    font-family: var(--sans);
    overflow-x: hidden;
    cursor: default;
  }

  /* ── SCANLINE OVERLAY ── */
  body::before {
    content: '';
    position: fixed;
    inset: 0;
    background: repeating-linear-gradient(0deg, transparent, transparent 2px, rgba(0,0,0,0.07) 2px, rgba(0,0,0,0.07) 4px);
    pointer-events: none;
    z-index: 9999;
  }

  /* ── GRID BACKGROUND ── */
  .grid-bg {
    position: fixed;
    inset: 0;
    background-image:
      linear-gradient(rgba(0,212,255,0.04) 1px, transparent 1px),
      linear-gradient(90deg, rgba(0,212,255,0.04) 1px, transparent 1px);
    background-size: 60px 60px;
    pointer-events: none;
    z-index: 0;
  }

  /* ── NAV ── */
  nav {
    position: fixed;
    top: 0; left: 0; right: 0;
    z-index: 100;
    background: rgba(4,12,20,0.92);
    backdrop-filter: blur(12px);
    border-bottom: 1px solid var(--border);
    padding: 0 2rem;
    display: flex;
    align-items: center;
    justify-content: space-between;
    height: 56px;
  }

  .nav-logo {
    font-family: var(--mono);
    font-size: 0.8rem;
    color: var(--cyan);
    letter-spacing: 2px;
    text-transform: uppercase;
  }

  .nav-links {
    display: flex;
    gap: 2rem;
    list-style: none;
  }

  .nav-links a {
    font-family: var(--mono);
    font-size: 0.72rem;
    color: var(--muted);
    text-decoration: none;
    letter-spacing: 1px;
    text-transform: uppercase;
    transition: color 0.2s;
  }

  .nav-links a:hover { color: var(--cyan); }

  /* ── SECTIONS ── */
  section {
    position: relative;
    z-index: 1;
    padding: 6rem 2rem;
    max-width: 1200px;
    margin: 0 auto;
  }

  /* ── HERO ── */
  #hero {
    min-height: 100vh;
    display: flex;
    flex-direction: column;
    justify-content: center;
    max-width: 100%;
    padding: 6rem 4rem;
    position: relative;
    overflow: hidden;
  }

  .hero-tag {
    font-family: var(--mono);
    font-size: 0.72rem;
    color: var(--cyan);
    letter-spacing: 4px;
    text-transform: uppercase;
    margin-bottom: 1.5rem;
    opacity: 0;
    animation: fadeUp 0.8s 0.2s forwards;
  }

  h1 {
    font-size: clamp(2.5rem, 6vw, 5.5rem);
    font-weight: 800;
    line-height: 1.05;
    letter-spacing: -2px;
    max-width: 900px;
    opacity: 0;
    animation: fadeUp 0.8s 0.4s forwards;
  }

  h1 span { color: var(--cyan); }

  .hero-sub {
    margin-top: 2rem;
    font-size: 1rem;
    color: var(--muted);
    max-width: 600px;
    line-height: 1.7;
    opacity: 0;
    animation: fadeUp 0.8s 0.6s forwards;
  }

  .hero-stats {
    display: flex;
    gap: 3rem;
    margin-top: 4rem;
    opacity: 0;
    animation: fadeUp 0.8s 0.8s forwards;
  }

  .stat-item { text-align: left; }

  .stat-num {
    font-family: var(--mono);
    font-size: 2.8rem;
    color: var(--cyan);
    line-height: 1;
  }

  .stat-label {
    font-size: 0.72rem;
    color: var(--muted);
    letter-spacing: 2px;
    text-transform: uppercase;
    margin-top: 0.3rem;
  }

  /* Glowing orb */
  .hero-orb {
    position: absolute;
    right: -100px;
    top: 50%;
    transform: translateY(-50%);
    width: 600px;
    height: 600px;
    border-radius: 50%;
    background: radial-gradient(circle, rgba(0,212,255,0.08) 0%, transparent 70%);
    animation: pulse 4s ease-in-out infinite;
  }

  @keyframes pulse {
    0%, 100% { transform: translateY(-50%) scale(1); opacity: 0.6; }
    50% { transform: translateY(-50%) scale(1.1); opacity: 1; }
  }

  /* ── SECTION HEADERS ── */
  .section-label {
    font-family: var(--mono);
    font-size: 0.68rem;
    color: var(--cyan);
    letter-spacing: 4px;
    text-transform: uppercase;
    margin-bottom: 0.75rem;
  }

  h2 {
    font-size: clamp(1.8rem, 4vw, 3rem);
    font-weight: 700;
    letter-spacing: -1px;
    margin-bottom: 0.5rem;
    line-height: 1.15;
  }

  .section-desc {
    color: var(--muted);
    font-size: 0.9rem;
    line-height: 1.7;
    max-width: 640px;
    margin-bottom: 3rem;
  }

  /* ── DIVIDER ── */
  .divider {
    border: none;
    border-top: 1px solid var(--border);
    margin: 0;
  }

  /* ── DATASET CARDS ── */
  .dataset-grid {
    display: grid;
    grid-template-columns: 1fr 1fr;
    gap: 1.5rem;
    margin-top: 2rem;
  }

  .card {
    background: var(--bg2);
    border: 1px solid var(--border);
    border-radius: 12px;
    padding: 2rem;
    position: relative;
    overflow: hidden;
    transition: border-color 0.3s, transform 0.3s;
  }

  .card:hover {
    border-color: rgba(0,212,255,0.4);
    transform: translateY(-2px);
  }

  .card::before {
    content: '';
    position: absolute;
    top: 0; left: 0; right: 0;
    height: 2px;
    background: linear-gradient(90deg, var(--cyan), transparent);
  }

  .card-title {
    font-family: var(--mono);
    font-size: 0.78rem;
    color: var(--cyan);
    letter-spacing: 2px;
    text-transform: uppercase;
    margin-bottom: 1.2rem;
  }

  .card h3 {
    font-size: 1.5rem;
    font-weight: 700;
    margin-bottom: 0.5rem;
  }

  .card p {
    font-size: 0.85rem;
    color: var(--muted);
    line-height: 1.6;
    margin-bottom: 1.5rem;
  }

  .dataset-stat {
    display: flex;
    justify-content: space-between;
    align-items: center;
    padding: 0.5rem 0;
    border-top: 1px solid var(--border);
    font-size: 0.82rem;
  }

  .dataset-stat span:first-child { color: var(--muted); }
  .dataset-stat span:last-child { font-family: var(--mono); color: var(--cyan); }

  /* ── TABS ── */
  .tabs {
    display: flex;
    gap: 0;
    border: 1px solid var(--border);
    border-radius: 8px;
    overflow: hidden;
    margin-bottom: 2rem;
    width: fit-content;
  }

  .tab-btn {
    background: transparent;
    border: none;
    color: var(--muted);
    font-family: var(--mono);
    font-size: 0.72rem;
    letter-spacing: 2px;
    text-transform: uppercase;
    padding: 0.7rem 1.5rem;
    cursor: pointer;
    transition: all 0.2s;
    border-right: 1px solid var(--border);
  }

  .tab-btn:last-child { border-right: none; }

  .tab-btn.active {
    background: rgba(0,212,255,0.1);
    color: var(--cyan);
  }

  .tab-btn:hover:not(.active) {
    background: rgba(0,212,255,0.05);
    color: var(--text);
  }

  .tab-content { display: none; }
  .tab-content.active { display: block; }

  /* ── CHART AREA ── */
  .chart-wrap {
    background: var(--bg2);
    border: 1px solid var(--border);
    border-radius: 12px;
    padding: 2rem;
    margin-bottom: 2rem;
  }

  .chart-title {
    font-family: var(--mono);
    font-size: 0.72rem;
    color: var(--muted);
    letter-spacing: 2px;
    text-transform: uppercase;
    margin-bottom: 1.5rem;
  }

  .bar-chart { display: flex; flex-direction: column; gap: 0.65rem; }

  .bar-row {
    display: grid;
    grid-template-columns: 130px 1fr 70px;
    align-items: center;
    gap: 0.75rem;
  }

  .bar-label {
    font-family: var(--mono);
    font-size: 0.72rem;
    color: var(--muted);
    text-align: right;
    white-space: nowrap;
  }

  .bar-track {
    height: 20px;
    background: rgba(255,255,255,0.04);
    border-radius: 4px;
    overflow: hidden;
  }

  .bar-fill {
    height: 100%;
    border-radius: 4px;
    width: 0;
    transition: width 1s cubic-bezier(.16,1,.3,1);
    position: relative;
  }

  .bar-fill::after {
    content: '';
    position: absolute;
    top: 0; right: 0; bottom: 0;
    width: 6px;
    background: rgba(255,255,255,0.3);
    border-radius: 0 4px 4px 0;
  }

  .bar-val {
    font-family: var(--mono);
    font-size: 0.72rem;
    color: var(--text);
  }

  /* ── DATA TABLE ── */
  .table-wrap {
    overflow-x: auto;
    border: 1px solid var(--border);
    border-radius: 12px;
  }

  table {
    width: 100%;
    border-collapse: collapse;
    font-size: 0.8rem;
  }

  thead th {
    background: var(--bg3);
    font-family: var(--mono);
    font-size: 0.65rem;
    letter-spacing: 2px;
    text-transform: uppercase;
    color: var(--cyan);
    padding: 0.9rem 1rem;
    text-align: left;
    border-bottom: 1px solid var(--border);
    white-space: nowrap;
  }

  tbody td {
    padding: 0.7rem 1rem;
    border-bottom: 1px solid var(--border);
    color: var(--text);
    font-family: var(--mono);
    font-size: 0.78rem;
    white-space: nowrap;
  }

  tbody tr:last-child td { border-bottom: none; }

  tbody tr:hover td { background: rgba(0,212,255,0.04); }

  .best { color: var(--green); font-weight: 700; }
  .worst { color: var(--red); }
  .mid { color: var(--yellow); }

  .badge {
    display: inline-block;
    padding: 0.15rem 0.5rem;
    border-radius: 3px;
    font-size: 0.65rem;
    letter-spacing: 1px;
    text-transform: uppercase;
  }

  .badge-ml { background: rgba(0,255,157,0.1); color: var(--green); border: 1px solid rgba(0,255,157,0.2); }
  .badge-dl { background: rgba(0,212,255,0.1); color: var(--cyan); border: 1px solid rgba(0,212,255,0.2); }
  .badge-smote { background: rgba(255,107,53,0.1); color: var(--orange); border: 1px solid rgba(255,107,53,0.2); }

  /* ── FINDING CARDS ── */
  .findings-grid {
    display: grid;
    grid-template-columns: repeat(3, 1fr);
    gap: 1.5rem;
    margin-top: 2rem;
  }

  .finding-card {
    background: var(--bg2);
    border: 1px solid var(--border);
    border-radius: 12px;
    padding: 1.75rem;
    position: relative;
    overflow: hidden;
    transition: all 0.3s;
  }

  .finding-card:hover {
    border-color: rgba(0,212,255,0.35);
    transform: translateY(-3px);
  }

  .finding-icon {
    font-size: 1.8rem;
    margin-bottom: 1rem;
  }

  .finding-card h4 {
    font-size: 1rem;
    font-weight: 700;
    margin-bottom: 0.6rem;
    color: var(--text);
  }

  .finding-card p {
    font-size: 0.82rem;
    color: var(--muted);
    line-height: 1.6;
  }

  /* ── COMPUTATIONAL COST ── */
  .compute-grid {
    display: grid;
    grid-template-columns: 1fr 1fr;
    gap: 1.5rem;
    margin-top: 2rem;
  }

  /* ── GUIDELINES TABLE ── */
  .guidelines-wrap {
    background: var(--bg2);
    border: 1px solid var(--border);
    border-radius: 12px;
    overflow: hidden;
    margin-top: 2rem;
  }

  .gl-row {
    display: grid;
    grid-template-columns: 2fr 1.2fr 1.2fr 2fr;
    gap: 0;
    border-bottom: 1px solid var(--border);
  }

  .gl-row:last-child { border-bottom: none; }

  .gl-header .gl-cell {
    background: var(--bg3);
    font-family: var(--mono);
    font-size: 0.65rem;
    color: var(--cyan);
    letter-spacing: 2px;
    text-transform: uppercase;
  }

  .gl-cell {
    padding: 1rem 1.2rem;
    font-size: 0.82rem;
    border-right: 1px solid var(--border);
    line-height: 1.5;
  }

  .gl-cell:last-child { border-right: none; }

  .gl-row:not(.gl-header):hover .gl-cell { background: rgba(0,212,255,0.03); }

  /* ── ANIMATIONS ── */
  @keyframes fadeUp {
    from { opacity: 0; transform: translateY(24px); }
    to { opacity: 1; transform: translateY(0); }
  }

  .reveal {
    opacity: 0;
    transform: translateY(30px);
    transition: opacity 0.7s ease, transform 0.7s ease;
  }

  .reveal.visible {
    opacity: 1;
    transform: translateY(0);
  }

  /* ── FOOTER ── */
  footer {
    position: relative;
    z-index: 1;
    border-top: 1px solid var(--border);
    padding: 2rem 4rem;
    display: flex;
    justify-content: space-between;
    align-items: center;
  }

  footer p {
    font-family: var(--mono);
    font-size: 0.68rem;
    color: var(--muted);
    letter-spacing: 1px;
  }

  .dot {
    display: inline-block;
    width: 6px; height: 6px;
    border-radius: 50%;
    background: var(--cyan);
    margin-right: 0.5rem;
    animation: blink 1.2s step-end infinite;
  }

  @keyframes blink {
    0%, 100% { opacity: 1; }
    50% { opacity: 0; }
  }

  @media (max-width: 768px) {
    #hero { padding: 6rem 2rem; }
    .hero-stats { flex-wrap: wrap; gap: 2rem; }
    .dataset-grid, .findings-grid, .compute-grid { grid-template-columns: 1fr; }
    .gl-row { grid-template-columns: 1fr 1fr; }
    nav { padding: 0 1rem; }
    .nav-links { display: none; }
    h1 { letter-spacing: -1px; }
  }
</style>
</head>
<body>

<div class="grid-bg"></div>

<!-- NAV -->
<nav>
  <div class="nav-logo"><span class="dot"></span>NIDS Survey 2024</div>
  <ul class="nav-links">
    <li><a href="#datasets">Datasets</a></li>
    <li><a href="#results">Results</a></li>
    <li><a href="#compute">Compute Cost</a></li>
    <li><a href="#findings">Findings</a></li>
    <li><a href="#guidelines">Guidelines</a></li>
  </ul>
</nav>

<!-- HERO -->
<div id="hero">
  <div class="hero-orb"></div>
  <div class="hero-tag">Conference Presentation · NIDS Survey</div>
  <h1>From <span>Data</span><br>to Defense</h1>
  <p class="hero-sub">An analytical overview of machine and deep learning models for Network Intrusion Detection Systems, evaluated across two benchmark datasets under a unified experimental framework.</p>
  <div class="hero-stats">
    <div class="stat-item">
      <div class="stat-num">7</div>
      <div class="stat-label">Models Evaluated</div>
    </div>
    <div class="stat-item">
      <div class="stat-num">2</div>
      <div class="stat-label">Benchmark Datasets</div>
    </div>
    <div class="stat-item">
      <div class="stat-num">8</div>
      <div class="stat-label">Evaluation Metrics</div>
    </div>
    <div class="stat-item">
      <div class="stat-num">4M+</div>
      <div class="stat-label">Records Processed</div>
    </div>
  </div>
</div>

<hr class="divider">

<!-- DATASETS -->
<section id="datasets">
  <div class="reveal">
    <div class="section-label">// Section 01</div>
    <h2>Benchmark Datasets</h2>
    <p class="section-desc">Two widely-used intrusion detection datasets selected to represent both classical and modern network traffic environments.</p>
  </div>

  <div class="dataset-grid reveal">
    <div class="card">
      <div class="card-title">Dataset 01</div>
      <h3>KDD Cup 99</h3>
      <p>One of the earliest and most widely-used IDS benchmarks. Contains diverse attack types across 41 features, enabling systematic ML evaluation despite its age.</p>
      <div class="dataset-stat"><span>Total Records</span><span>~4,000,000</span></div>
      <div class="dataset-stat"><span>Features</span><span>41</span></div>
      <div class="dataset-stat"><span>Attack Types</span><span>10 (after filtering)</span></div>
      <div class="dataset-stat"><span>Largest Class</span><span>smurf · 2,807,886</span></div>
      <div class="dataset-stat"><span>Normal Traffic</span><span>972,781</span></div>
    </div>

    <div class="card">
      <div class="card-title">Dataset 02</div>
      <h3>UNSW-NB15</h3>
      <p>A modern dataset with realistic traffic patterns and diverse attack categories. Reflects contemporary network environments with significant class imbalance.</p>
      <div class="dataset-stat"><span>Total Records</span><span>~2,367,624</span></div>
      <div class="dataset-stat"><span>Attack Categories</span><span>9</span></div>
      <div class="dataset-stat"><span>Benign Traffic</span><span>2,237,731</span></div>
      <div class="dataset-stat"><span>Rarest Class</span><span>Worms · 158</span></div>
      <div class="dataset-stat"><span>Imbalance Ratio</span><span>High</span></div>
    </div>
  </div>
</section>

<hr class="divider">

<!-- RESULTS -->
<section id="results">
  <div class="reveal">
    <div class="section-label">// Section 02</div>
    <h2>Model Performance Results</h2>
    <p class="section-desc">Comprehensive metrics including balanced accuracy, F1-macro, G-Mean, precision, and recall evaluated across all models on both datasets — with and without SMOTE oversampling.</p>
  </div>

  <div class="reveal">
    <div class="tabs">
      <button class="tab-btn active" onclick="switchTab(event,'unsw-raw')">UNSW · Raw</button>
      <button class="tab-btn" onclick="switchTab(event,'unsw-smote')">UNSW · SMOTE</button>
      <button class="tab-btn" onclick="switchTab(event,'kdd-raw')">KDD Cup · Raw</button>
      <button class="tab-btn" onclick="switchTab(event,'kdd-smote')">KDD Cup · SMOTE</button>
    </div>

    <!-- UNSW RAW -->
    <div id="unsw-raw" class="tab-content active">
      <div class="chart-wrap">
        <div class="chart-title">// Balanced Accuracy — UNSW-NB15 (Raw)</div>
        <div class="bar-chart" id="chart-unsw-raw"></div>
      </div>
      <div class="table-wrap">
        <table>
          <thead>
            <tr>
              <th>Model</th><th>Type</th><th>Accuracy</th><th>Balanced Acc.</th><th>Precision</th><th>Recall</th><th>F1-Macro</th><th>G-Mean</th>
            </tr>
          </thead>
          <tbody>
            <tr><td>Random Forest</td><td><span class="badge badge-ml">ML</span></td><td>0.9757</td><td class="best">0.5657</td><td class="best">0.6197</td><td class="best">0.5657</td><td class="best">0.5810</td><td class="best">0.4822</td></tr>
            <tr><td>XGBoost</td><td><span class="badge badge-ml">ML</span></td><td>0.9738</td><td>0.4247</td><td>0.6435</td><td>0.4247</td><td>0.4510</td><td class="worst">0.0180</td></tr>
            <tr><td>Decision Tree</td><td><span class="badge badge-ml">ML</span></td><td>0.9753</td><td>0.4703</td><td>0.5760</td><td>0.4703</td><td>0.4850</td><td>0.0238</td></tr>
            <tr><td>LSTM-CNN</td><td><span class="badge badge-dl">DL</span></td><td>0.9729</td><td>0.3823</td><td>0.4673</td><td>0.3823</td><td>0.3950</td><td>0.0001</td></tr>
            <tr><td>ANN</td><td><span class="badge badge-dl">DL</span></td><td>0.9589</td><td>0.3744</td><td>0.3568</td><td>0.3744</td><td>0.2990</td><td class="worst">≈0</td></tr>
            <tr><td>LSTM</td><td><span class="badge badge-dl">DL</span></td><td>0.9673</td><td>0.3287</td><td>0.3546</td><td>0.3287</td><td>0.3250</td><td class="worst">≈0</td></tr>
            <tr><td>CNN</td><td><span class="badge badge-dl">DL</span></td><td class="worst">0.9524</td><td class="worst">0.2781</td><td class="worst">0.3386</td><td class="worst">0.2781</td><td class="worst">0.2730</td><td class="worst">≈0</td></tr>
          </tbody>
        </table>
      </div>
    </div>

    <!-- UNSW SMOTE -->
    <div id="unsw-smote" class="tab-content">
      <div class="chart-wrap">
        <div class="chart-title">// Balanced Accuracy — UNSW-NB15 (After SMOTE) — Degradation observed</div>
        <div class="bar-chart" id="chart-unsw-smote"></div>
      </div>
      <div class="table-wrap">
        <table>
          <thead>
            <tr>
              <th>Model</th><th>Accuracy</th><th>Balanced Acc.</th><th>Precision</th><th>Recall</th><th>F1-Macro</th><th>G-Mean</th>
            </tr>
          </thead>
          <tbody>
            <tr><td>LSTM-CNN <span class="badge badge-smote">SMOTE</span></td><td class="worst">0.0112</td><td>0.1241</td><td>0.0918</td><td>0.1241</td><td class="worst">0.0077</td><td>0.0017</td></tr>
            <tr><td>Random Forest <span class="badge badge-smote">SMOTE</span></td><td>0.9194</td><td class="best">0.1001</td><td class="best">0.1919</td><td class="best">0.1001</td><td class="best">0.0961</td><td class="worst">≈0</td></tr>
            <tr><td>ANN <span class="badge badge-smote">SMOTE</span></td><td>0.9194</td><td>0.1000</td><td>0.0919</td><td>0.1000</td><td>0.0958</td><td class="worst">≈0</td></tr>
            <tr><td>CNN <span class="badge badge-smote">SMOTE</span></td><td>0.9194</td><td>0.1000</td><td>0.0919</td><td>0.1000</td><td>0.0958</td><td class="worst">≈0</td></tr>
            <tr><td>XGBoost <span class="badge badge-smote">SMOTE</span></td><td>0.1104</td><td>0.1108</td><td>0.1044</td><td>0.1108</td><td>0.0248</td><td class="worst">≈0</td></tr>
            <tr><td>Decision Tree <span class="badge badge-smote">SMOTE</span></td><td>0.8633</td><td>0.0960</td><td>0.0930</td><td>0.0960</td><td>0.0937</td><td class="worst">≈0</td></tr>
            <tr><td>LSTM <span class="badge badge-smote">SMOTE</span></td><td class="worst">0.0352</td><td class="worst">0.0683</td><td class="worst">0.0814</td><td class="worst">0.0683</td><td class="worst">0.0097</td><td class="worst">≈0</td></tr>
          </tbody>
        </table>
      </div>
      <div class="chart-wrap" style="margin-top:1.5rem; border-color: rgba(255,61,90,0.2)">
        <div class="chart-title" style="color:var(--red)">// ⚠ SMOTE degraded ALL metrics on UNSW-NB15 — Not recommended for high-dimensional imbalanced data</div>
        <p style="font-size:0.85rem; color:var(--muted); line-height:1.7">Applying SMOTE to the UNSW-NB15 dataset caused substantial deterioration across all models. Balanced accuracy, F1-macro, and G-Mean dropped significantly compared to raw data settings, indicating that sample-level oversampling is ineffective for high-dimensional complex datasets. Model robustness and algorithm-level imbalance handling are more critical.</p>
      </div>
    </div>

    <!-- KDD RAW -->
    <div id="kdd-raw" class="tab-content">
      <div class="chart-wrap">
        <div class="chart-title">// Balanced Accuracy — KDD Cup 99 (Raw)</div>
        <div class="bar-chart" id="chart-kdd-raw"></div>
      </div>
      <div class="table-wrap">
        <table>
          <thead>
            <tr>
              <th>Model</th><th>Type</th><th>Accuracy</th><th>Balanced Acc.</th><th>Precision</th><th>Recall</th><th>F1-Macro</th><th>G-Mean</th>
            </tr>
          </thead>
          <tbody>
            <tr><td>Random Forest</td><td><span class="badge badge-ml">ML</span></td><td class="best">0.9999</td><td class="best">0.9940</td><td class="best">0.9969</td><td class="best">0.9940</td><td class="best">0.9955</td><td class="best">0.9939</td></tr>
            <tr><td>XGBoost</td><td><span class="badge badge-ml">ML</span></td><td>0.9997</td><td>0.9909</td><td>0.9941</td><td>0.9909</td><td>0.9925</td><td>0.9908</td></tr>
            <tr><td>LSTM-CNN</td><td><span class="badge badge-dl">DL</span></td><td>0.9992</td><td>0.9848</td><td>0.9687</td><td>0.9848</td><td>0.9758</td><td>0.9846</td></tr>
            <tr><td>Decision Tree</td><td><span class="badge badge-ml">ML</span></td><td>0.9988</td><td>0.9614</td><td>0.9659</td><td>0.9614</td><td>0.9622</td><td>0.9586</td></tr>
            <tr><td>LSTM</td><td><span class="badge badge-dl">DL</span></td><td>0.9985</td><td>0.9673</td><td>0.9401</td><td>0.9673</td><td>0.9517</td><td>0.9663</td></tr>
            <tr><td>ANN</td><td><span class="badge badge-dl">DL</span></td><td>0.9987</td><td>0.9427</td><td>0.9622</td><td>0.9427</td><td>0.9519</td><td>0.9362</td></tr>
            <tr><td>CNN</td><td><span class="badge badge-dl">DL</span></td><td>0.9899</td><td class="worst">0.6256</td><td>0.7080</td><td class="worst">0.6256</td><td>0.6446</td><td class="worst">0.0051</td></tr>
          </tbody>
        </table>
      </div>
    </div>

    <!-- KDD SMOTE -->
    <div id="kdd-smote" class="tab-content">
      <div class="chart-wrap">
        <div class="chart-title">// Balanced Accuracy — KDD Cup 99 (After SMOTE) — Selective improvements</div>
        <div class="bar-chart" id="chart-kdd-smote"></div>
      </div>
      <div class="table-wrap">
        <table>
          <thead>
            <tr>
              <th>Model</th><th>Accuracy</th><th>Balanced Acc.</th><th>Precision</th><th>Recall</th><th>F1-Macro</th><th>G-Mean</th>
            </tr>
          </thead>
          <tbody>
            <tr><td>Random Forest <span class="badge badge-smote">SMOTE</span></td><td class="best">0.9999</td><td>0.9922</td><td class="best">0.9978</td><td>0.9922</td><td class="best">0.9949</td><td>0.9920</td></tr>
            <tr><td>XGBoost <span class="badge badge-smote">SMOTE</span></td><td>0.9997</td><td class="best">0.9971</td><td>0.9864</td><td class="best">0.9971</td><td>0.9916</td><td class="best">0.9971</td></tr>
            <tr><td>ANN <span class="badge badge-smote">SMOTE</span></td><td>0.9926</td><td>0.9953</td><td>0.7993</td><td>0.9954</td><td>0.8655</td><td>0.9953</td></tr>
            <tr><td>LSTM-CNN <span class="badge badge-smote">SMOTE</span></td><td>0.9992</td><td>0.9907</td><td>0.9589</td><td>0.9908</td><td>0.9737</td><td>0.9906</td></tr>
            <tr><td>Decision Tree <span class="badge badge-smote">SMOTE</span></td><td>0.9970</td><td>0.9931</td><td>0.9252</td><td>0.9931</td><td>0.9386</td><td>0.9930</td></tr>
            <tr><td>LSTM <span class="badge badge-smote">SMOTE</span></td><td>0.9981</td><td>0.9980</td><td>0.9111</td><td>0.9912</td><td>0.9431</td><td>0.9910</td></tr>
            <tr><td>CNN <span class="badge badge-smote">SMOTE</span></td><td>0.9494</td><td>0.9721</td><td class="worst">0.5231</td><td>0.9721</td><td class="worst">0.6199</td><td>0.9717</td></tr>
          </tbody>
        </table>
      </div>
      <div class="chart-wrap" style="margin-top:1.5rem; border-color: rgba(0,255,157,0.2)">
        <div class="chart-title" style="color:var(--green)">// ✓ SMOTE on KDD Cup 99 — Improvements for DL models, marginal effect on tree-based</div>
        <p style="font-size:0.85rem; color:var(--muted); line-height:1.7">SMOTE improved minority class recall for deep learning models notably (ANN: 0.94 → 0.99, CNN: 0.63 → 0.97). However, precision dropped, indicating more false positives. Tree-based models (RF, XGBoost) were largely unaffected — they inherently handle class imbalance well and do not benefit significantly from data-level oversampling.</p>
      </div>
    </div>
  </div>
</section>

<hr class="divider">

<!-- COMPUTE COST -->
<section id="compute">
  <div class="reveal">
    <div class="section-label">// Section 03</div>
    <h2>Computational Cost Analysis</h2>
    <p class="section-desc">Inference time and memory footprint are critical factors for real-world NIDS deployment. Deep learning architectures show significantly higher latency and resource consumption.</p>
  </div>

  <div class="compute-grid reveal">
    <div class="chart-wrap">
      <div class="chart-title">// Inference Time (seconds) — UNSW-NB15</div>
      <div class="bar-chart" id="chart-time-unsw"></div>
    </div>
    <div class="chart-wrap">
      <div class="chart-title">// Inference Time (seconds) — KDD Cup 99</div>
      <div class="bar-chart" id="chart-time-kdd"></div>
    </div>
  </div>

  <div class="compute-grid reveal">
    <div class="chart-wrap">
      <div class="chart-title">// Memory Usage (MB) — UNSW-NB15 · Note: RF 22,424 MB</div>
      <div class="bar-chart" id="chart-mem-unsw"></div>
    </div>
    <div class="chart-wrap">
      <div class="chart-title">// Memory Usage (MB) — KDD Cup 99</div>
      <div class="bar-chart" id="chart-mem-kdd"></div>
    </div>
  </div>

  <div class="table-wrap reveal" style="margin-top:1.5rem">
    <table>
      <thead>
        <tr><th>Model</th><th>Type</th><th>UNSW Infer. Time (s)</th><th>UNSW Memory (MB)</th><th>KDD Infer. Time (s)</th><th>KDD Memory (MB)</th></tr>
      </thead>
      <tbody>
        <tr><td>Decision Tree</td><td><span class="badge badge-ml">ML</span></td><td class="best">0.025</td><td>1,579</td><td class="best">0.026</td><td>4,959</td></tr>
        <tr><td>XGBoost</td><td><span class="badge badge-ml">ML</span></td><td>0.333</td><td class="best">1,571</td><td>0.226</td><td class="best">4,959</td></tr>
        <tr><td>Random Forest</td><td><span class="badge badge-ml">ML</span></td><td>0.571</td><td class="worst">22,424</td><td>0.373</td><td>4,989</td></tr>
        <tr><td>ANN</td><td><span class="badge badge-dl">DL</span></td><td>1.622</td><td>5,267</td><td>1.410</td><td>4,503</td></tr>
        <tr><td>CNN</td><td><span class="badge badge-dl">DL</span></td><td>1.649</td><td>4,375</td><td>1.314</td><td>5,096</td></tr>
        <tr><td>LSTM</td><td><span class="badge badge-dl">DL</span></td><td>1.902</td><td>4,710</td><td>1.475</td><td>5,104</td></tr>
        <tr><td>LSTM-CNN</td><td><span class="badge badge-dl">DL</span></td><td class="worst">3.721</td><td>4,720</td><td class="worst">2.995</td><td>5,112</td></tr>
      </tbody>
    </table>
  </div>
</section>

<hr class="divider">

<!-- KEY FINDINGS -->
<section id="findings">
  <div class="reveal">
    <div class="section-label">// Section 04</div>
    <h2>Key Findings</h2>
    <p class="section-desc">Six critical insights derived from unified experimental evaluation across models and datasets.</p>
  </div>

  <div class="findings-grid">
    <div class="finding-card reveal">
      <div class="finding-icon">🌲</div>
      <h4>Tree Models Dominate on Structured Data</h4>
      <p>Random Forest and XGBoost achieved near-perfect F1-macro (>0.99) on KDD Cup 99. Ensemble methods consistently outperformed all other approaches on structured, tabular datasets.</p>
    </div>
    <div class="finding-card reveal">
      <div class="finding-icon">⚖️</div>
      <h4>Accuracy is a Misleading Metric</h4>
      <p>Multiple DL models reported >96% accuracy on UNSW-NB15, yet their balanced accuracy and G-Mean revealed near-zero minority class detection, exposing the danger of relying on a single metric.</p>
    </div>
    <div class="finding-card reveal">
      <div class="finding-icon">🧠</div>
      <h4>Deep Learning: High Cost, Mixed Returns</h4>
      <p>CNN-LSTM hybrid achieved strong KDD performance (F1: 0.976) but requires 3.7s inference and >5GB memory. The computational cost limits real-time applicability significantly.</p>
    </div>
    <div class="finding-card reveal">
      <div class="finding-icon">🚫</div>
      <h4>SMOTE Harmful on High-Dimensional Data</h4>
      <p>Applying SMOTE to UNSW-NB15 degraded all metrics across all models substantially. Simple sample-level balancing cannot compensate for feature complexity and distribution shifts.</p>
    </div>
    <div class="finding-card reveal">
      <div class="finding-icon">✅</div>
      <h4>SMOTE Helps DL on KDD Cup 99</h4>
      <p>ANN balanced accuracy improved from 0.943 to 0.995, and CNN from 0.626 to 0.972 after SMOTE on KDD Cup 99. However, this came at the cost of precision degradation.</p>
    </div>
    <div class="finding-card reveal">
      <div class="finding-icon">⚡</div>
      <h4>Decision Tree Best for Real-Time</h4>
      <p>With ~0.025s inference time and the smallest memory footprint, Decision Trees are optimal for latency-sensitive deployments where sub-second detection is a hard requirement.</p>
    </div>
  </div>
</section>

<hr class="divider">

<!-- GUIDELINES -->
<section id="guidelines">
  <div class="reveal">
    <div class="section-label">// Section 05</div>
    <h2>Dataset–Model Selection Guidelines</h2>
    <p class="section-desc">Practical decision framework for selecting the right model and strategy based on deployment requirements and dataset characteristics.</p>
  </div>

  <div class="guidelines-wrap reveal">
    <div class="gl-row gl-header">
      <div class="gl-cell">Deployment Scenario</div>
      <div class="gl-cell">Recommended Model</div>
      <div class="gl-cell">Dataset</div>
      <div class="gl-cell">Key Evidence</div>
    </div>
    <div class="gl-row">
      <div class="gl-cell">Structured, low-complexity traffic</div>
      <div class="gl-cell"><span class="badge badge-ml">RF / XGBoost</span></div>
      <div class="gl-cell" style="font-family:var(--mono); font-size:0.78rem; color:var(--cyan)">KDD Cup 99</div>
      <div class="gl-cell" style="color:var(--muted)">F1-macro &gt;0.99; near-perfect balanced accuracy</div>
    </div>
    <div class="gl-row">
      <div class="gl-cell">High-dimensional, imbalanced traffic</div>
      <div class="gl-cell"><span class="badge badge-ml">RF</span></div>
      <div class="gl-cell" style="font-family:var(--mono); font-size:0.78rem; color:var(--cyan)">UNSW-NB15</div>
      <div class="gl-cell" style="color:var(--muted)">Best balanced accuracy (0.57) and G-Mean (0.48) among all tested models</div>
    </div>
    <div class="gl-row">
      <div class="gl-cell">Real-time / latency-sensitive deployment</div>
      <div class="gl-cell"><span class="badge badge-ml">DT</span></div>
      <div class="gl-cell" style="font-family:var(--mono); font-size:0.78rem; color:var(--cyan)">Both</div>
      <div class="gl-cell" style="color:var(--muted)">Fastest inference (~0.02s); lowest memory footprint across all datasets</div>
    </div>
    <div class="gl-row">
      <div class="gl-cell">Sequential / temporal attack patterns</div>
      <div class="gl-cell"><span class="badge badge-dl">LSTM / CNN-LSTM</span></div>
      <div class="gl-cell" style="font-family:var(--mono); font-size:0.78rem; color:var(--cyan)">KDD Cup 99</div>
      <div class="gl-cell" style="color:var(--muted)">Strong recall on ordered flow attacks; temporal dependency modeling</div>
    </div>
    <div class="gl-row">
      <div class="gl-cell">Imbalanced data + deep learning</div>
      <div class="gl-cell"><span class="badge badge-smote">SMOTE + ANN/LSTM</span></div>
      <div class="gl-cell" style="font-family:var(--mono); font-size:0.78rem; color:var(--cyan)">KDD Cup 99</div>
      <div class="gl-cell" style="color:var(--muted)">Balanced accuracy improved from 0.94 to 0.99; recall gains outweigh precision drop</div>
    </div>
    <div class="gl-row">
      <div class="gl-cell">High-dimensional data + imbalance</div>
      <div class="gl-cell"><span class="badge badge-ml">RF (no SMOTE)</span></div>
      <div class="gl-cell" style="font-family:var(--mono); font-size:0.78rem; color:var(--cyan)">UNSW-NB15</div>
      <div class="gl-cell" style="color:var(--muted)">SMOTE degraded all metrics on high-dimensional data; avoid sample-level balancing</div>
    </div>
  </div>

  <!-- Conclusion box -->
  <div class="card reveal" style="margin-top:2rem; border-color:rgba(0,255,157,0.2)">
    <div class="card-title" style="color:var(--green)">// Conclusion & Future Directions</div>
    <h3 style="margin-bottom:1rem; font-size:1.2rem">Ensemble Methods Offer the Best Production Trade-off</h3>
    <p>Decision Trees achieve inference times as low as 0.02s for real-time monitoring, while hybrid CNN-LSTM models exceed 3s latency with >5GB memory — suitable only for offline analysis. RF and XGBoost provide the optimal balance of strong detection performance with manageable inference costs, making them the most practical choice for production NIDS environments.</p>
    <p style="margin-top:1rem">Future research should explore <strong style="color:var(--cyan)">transformer-based architectures</strong> for NIDS — leveraging self-attention for parallel processing and better capture of global traffic patterns, potentially overcoming the computational bottlenecks observed in LSTM and CNN-LSTM models.</p>
  </div>
</section>

<footer>
  <p>From Data to Defense · NIDS Survey · Evaluation of ML & DL for Network Intrusion Detection</p>
  <p style="color:var(--cyan)">KDD Cup 99 · UNSW-NB15 · 7 Models · Unified Evaluation Framework</p>
</footer>

<script>
// ── BAR CHART RENDERER ──
function renderBar(containerId, data, colorFn) {
  const el = document.getElementById(containerId);
  if (!el) return;
  const max = Math.max(...data.map(d => d.val));
  el.innerHTML = '';
  data.forEach((d, i) => {
    const pct = (d.val / max * 100).toFixed(1);
    const color = colorFn ? colorFn(d, i) : `hsl(${180 + i * 15}, 70%, 55%)`;
    const row = document.createElement('div');
    row.className = 'bar-row';
    row.innerHTML = `
      <div class="bar-label">${d.label}</div>
      <div class="bar-track"><div class="bar-fill" data-width="${pct}" style="background:${color}; width:0%"></div></div>
      <div class="bar-val">${d.val.toFixed ? d.val.toFixed(4) : d.val}</div>`;
    el.appendChild(row);
  });
  requestAnimationFrame(() => {
    el.querySelectorAll('.bar-fill').forEach(b => b.style.width = b.dataset.width + '%');
  });
}

const ml_color = () => 'linear-gradient(90deg, #00d4ff, #00a8cc)';
const dl_color = () => 'linear-gradient(90deg, #8b5cf6, #6d28d9)';
const smote_color = () => 'linear-gradient(90deg, #ff6b35, #cc4400)';

function modelColor(d) {
  if (['RF','DT','XGBoost'].includes(d.type)) return 'linear-gradient(90deg, #00d4ff, #00a8cc)';
  return 'linear-gradient(90deg, #8b5cf6, #6d28d9)';
}

// UNSW RAW balanced acc
renderBar('chart-unsw-raw', [
  {label:'Random Forest', val:0.5657, type:'RF'},
  {label:'Decision Tree', val:0.4703, type:'DT'},
  {label:'XGBoost',       val:0.4247, type:'XGBoost'},
  {label:'LSTM-CNN',      val:0.3823, type:'dl'},
  {label:'ANN',           val:0.3744, type:'dl'},
  {label:'LSTM',          val:0.3287, type:'dl'},
  {label:'CNN',           val:0.2781, type:'dl'},
], modelColor);

// UNSW SMOTE balanced acc
renderBar('chart-unsw-smote', [
  {label:'LSTM-CNN',      val:0.1241, type:'dl'},
  {label:'XGBoost',       val:0.1108, type:'XGBoost'},
  {label:'Random Forest', val:0.1001, type:'RF'},
  {label:'ANN',           val:0.1000, type:'dl'},
  {label:'CNN',           val:0.1000, type:'dl'},
  {label:'DT',            val:0.0960, type:'DT'},
  {label:'LSTM',          val:0.0683, type:'dl'},
], (d) => smote_color());

// KDD RAW balanced acc
renderBar('chart-kdd-raw', [
  {label:'Random Forest', val:0.9940, type:'RF'},
  {label:'XGBoost',       val:0.9909, type:'XGBoost'},
  {label:'LSTM-CNN',      val:0.9848, type:'dl'},
  {label:'LSTM',          val:0.9673, type:'dl'},
  {label:'Decision Tree', val:0.9614, type:'DT'},
  {label:'ANN',           val:0.9427, type:'dl'},
  {label:'CNN',           val:0.6256, type:'dl'},
], modelColor);

// KDD SMOTE balanced acc
renderBar('chart-kdd-smote', [
  {label:'LSTM',          val:0.9980, type:'dl'},
  {label:'ANN',           val:0.9953, type:'dl'},
  {label:'XGBoost',       val:0.9971, type:'XGBoost'},
  {label:'DT',            val:0.9931, type:'DT'},
  {label:'Random Forest', val:0.9922, type:'RF'},
  {label:'LSTM-CNN',      val:0.9907, type:'dl'},
  {label:'CNN',           val:0.9721, type:'dl'},
], (d) => smote_color());

// INFERENCE TIME UNSW
renderBar('chart-time-unsw', [
  {label:'DT',       val:0.025, type:'DT'},
  {label:'XGBoost',  val:0.333, type:'XGBoost'},
  {label:'RF',       val:0.571, type:'RF'},
  {label:'ANN',      val:1.622, type:'dl'},
  {label:'CNN',      val:1.649, type:'dl'},
  {label:'LSTM',     val:1.902, type:'dl'},
  {label:'LSTM-CNN', val:3.721, type:'dl'},
], (d) => d.type === 'dl' ? 'linear-gradient(90deg,#ff3d5a,#cc2a45)' : 'linear-gradient(90deg,#00d4ff,#00a8cc)');

// INFERENCE TIME KDD
renderBar('chart-time-kdd', [
  {label:'DT',       val:0.026, type:'DT'},
  {label:'XGBoost',  val:0.226, type:'XGBoost'},
  {label:'RF',       val:0.373, type:'RF'},
  {label:'CNN',      val:1.314, type:'dl'},
  {label:'ANN',      val:1.410, type:'dl'},
  {label:'LSTM',     val:1.475, type:'dl'},
  {label:'LSTM-CNN', val:2.995, type:'dl'},
], (d) => d.type === 'dl' ? 'linear-gradient(90deg,#ff3d5a,#cc2a45)' : 'linear-gradient(90deg,#00d4ff,#00a8cc)');

// MEMORY UNSW (cap RF for display)
renderBar('chart-mem-unsw', [
  {label:'XGBoost',  val:1571,  type:'XGBoost'},
  {label:'DT',       val:1580,  type:'DT'},
  {label:'CNN',      val:4375,  type:'dl'},
  {label:'LSTM',     val:4710,  type:'dl'},
  {label:'LSTM-CNN', val:4720,  type:'dl'},
  {label:'ANN',      val:5267,  type:'dl'},
  {label:'RF ⚠',    val:22424, type:'RF'},
], (d) => d.label.includes('RF') ? 'linear-gradient(90deg,#ff6b35,#cc4400)' : d.type === 'dl' ? 'linear-gradient(90deg,#8b5cf6,#6d28d9)' : 'linear-gradient(90deg,#00d4ff,#00a8cc)');

// MEMORY KDD
renderBar('chart-mem-kdd', [
  {label:'ANN',      val:4503, type:'dl'},
  {label:'DT',       val:4959, type:'DT'},
  {label:'XGBoost',  val:4959, type:'XGBoost'},
  {label:'RF',       val:4989, type:'RF'},
  {label:'CNN',      val:5096, type:'dl'},
  {label:'LSTM',     val:5104, type:'dl'},
  {label:'LSTM-CNN', val:5112, type:'dl'},
], (d) => d.type === 'dl' ? 'linear-gradient(90deg,#8b5cf6,#6d28d9)' : 'linear-gradient(90deg,#00d4ff,#00a8cc)');

// ── TAB SWITCH ──
function switchTab(e, id) {
  document.querySelectorAll('.tab-btn').forEach(b => b.classList.remove('active'));
  document.querySelectorAll('.tab-content').forEach(c => c.classList.remove('active'));
  e.target.classList.add('active');
  document.getElementById(id).classList.add('active');
  // Re-trigger bar animations
  setTimeout(() => {
    document.querySelectorAll('#' + id + ' .bar-fill').forEach(b => {
      const w = b.dataset.width;
      b.style.width = '0%';
      requestAnimationFrame(() => setTimeout(() => b.style.width = w + '%', 50));
    });
  }, 50);
}

// ── SCROLL REVEAL ──
const observer = new IntersectionObserver((entries) => {
  entries.forEach(e => { if (e.isIntersecting) { e.target.classList.add('visible'); } });
}, { threshold: 0.1 });

document.querySelectorAll('.reveal').forEach(el => observer.observe(el));

// ── ANIMATE BARS ON LOAD ──
setTimeout(() => {
  document.querySelectorAll('#unsw-raw .bar-fill').forEach(b => b.style.width = b.dataset.width + '%');
}, 300);
</script>
</body>
</html>