King-8/academic-history-dataset

Academic History Dataset — Predicting Semester 1 Grades (Combined Version)

This dataset contains academic history from 9th to 12th grade, organized by class and semester.
It is designed to train a regression model to predict Semester 1 grades (as GPA values) for 12th grade classes based on past performance.

---

Dataset Structure

The dataset is stored as a single CSV file:

• academic_history_combined.csv — Contains both training rows (9th–11th grade) and prediction rows (12th grade Semester 1) with a split column to distinguish them.

Each row corresponds to one class in one semester.

---

Columns and Their Meaning

Column	Type	Description
year	String	School year (e.g., 2022–2023).
grade_level	Integer	Grade level (9–12).
semester	Integer	Semester number (1 or 2).
subject_type	Categorical	Broad course type (Core, Elective, Honors, Dual Enrollment).
subject	Categorical	Broad subject area (English, Science, Math, etc.).
class_name	Categorical	Specific course name (Bio 1 Hon, AP Stats, Band 2, etc.).
difficulty_level	Integer	Course rigor (1 = Regular/Elective, 2 = Honors, 3 = AP/Dual Enrollment).
cumulative_unweighted_GPA	Float	Cumulative unweighted GPA up to that semester.
cumulative_weighted_GPA	Float	Cumulative weighted GPA up to that semester.
daily_attendance	Float	Percentage of days attended for that class/semester.
study_hours	Float	Hours studied per week (planned or estimated). May be blank.
test_name	Categorical	Name of a major test or exam (S1 Exam, EOC, FSA ELA, etc.). May be blank.
test_score	Float	Percentage score on that test/exam. May be blank.
letter_grade	Categorical	Letter grade for the semester (A, B+, etc.) — filled for 9th–11th grade, blank for 12th grade.
GPA_value	Float	Target column: numeric GPA value of the semester grade (A=4.0, B+=3.3, etc.) — filled for 9th–11th, blank for 12th grade.
Predicted_GPA_Value	Float	Model-predicted GPA value for 12th grade Semester 1 rows.
Predicted_Letter_Grade	Categorical	Model-predicted letter grade for 12th grade Semester 1 rows.
split	Categorical	Indicates whether a row is part of the train set (9th–11th grade) or the predict set (12th grade Semester 1).

---

Splits

This dataset uses a split column instead of separate files:

• split = "train" rows correspond to 9th–11th grade training data with GPA_value filled.
• split = "predict" rows correspond to 12th grade Semester 1 rows with model-predicted GPA and letter grade.

You can easily filter by split in Python or on the Dataset Viewer.

Example:

from datasets import load_dataset

dataset = load_dataset("King-8/academic-history-dataset")

# Convert to pandas
df = dataset['train'].to_pandas()

train_rows = df[df['split'] == 'train']
predict_rows = df[df['split'] == 'predict']
Goal
The goal is to train a regression model that learns from 9th–11th grade data (with GPA_value filled) to predict GPA_value for 12th grade Semester 1 classes.

Input features: Subject, Difficulty Level, Cumulative GPAs, Attendance, Test Scores, Study Hours.

Target: GPA_value (numeric representation of semester grade).

Predicted GPA_values can be converted back into letter grades using the provided scale.

Letter Grade Conversion
GPA_Value	Letter Grade
≥3.85	A
3.7–3.84	A-
3.3–3.69	B+
3.0–3.29	B
2.7–2.99	B-
2.3–2.69	C+
2.0–2.29	C
1.7–1.99	C-
1.3–1.69	D+
1.0–1.29	D
<1.0	F

Usage
Train a regression model on rows where split = "train" using GPA_value as the target.

Use the trained model to predict GPA for rows where split = "predict".

Convert predicted GPA_value to letter grades using the provided scale.

---