Instruction file imported from yancheesetofer/final-skripsi (
.cursor/rules/results_reference.mdc). Copyright stays with the author.
Map of Key Result Files
When writing Bab 4 (Evaluasi dan Analisis), refer to these files as the source of truth.
Performance Metrics & Comparison
- File:
results_comparison_800_vs_90.md&results/report_*.txt - Content: Quantitative evidence of model performance (Accuracy, Precision, Recall, F1). Shows the superiority of the 800-sample model over the 90-sample model.
Feature & Behavior Analysis
- File:
results/feature_importance_Random Forest.png - Content: A bar plot showing the most influential features for detecting cheating.
- File:
detection_results/cheater_vs_normal_comparison.png - Content: Visual comparison of aggregate behavior between detected cheaters and normal users.
Case Studies (Studi Kasus)
- Files:
detection_results/offender_analysis_user_*.png - Content: Individual evidence plots for specific users flagged as suspicious. Use these as examples in the qualitative analysis section.
Data & Models
- File:
data/processed_artificial_features_V2.csv- The final dataset used for training. - File:
results/model_Random Forest.joblib- The final, saved, best-performing model. - File:
FINAL_SUMMARY_800_SAMPLES.md- High-level summary of the final experiment.
