List your company
ResearchFormerly WizWand

Adversarial Attack on Kaggle Credit Card Fraud

99Average Cumulative RewardBest Baseline
31.448.9566.584.05Feb 4, 2025
Updated 7mo ago

Evaluation results

MethodLinks
Best BaselineFeatures Fixed (%)=0%, Features Unknown (%)=0%, Victim Classifier=Random Forest2025.0299
Best BaselineFeatures Fixed (%)=25%, Features Unknown (%)=0%, Victim Classifier=Random Forest2025.0292
FRAUD-RLAFeatures Fixed (%)=0%, Features Unknown (%)=0%, FRAUD-RLA Time (Number of frauds)=4000, Victim Classifier=Random Forest2025.0291
Best BaselineFeatures Fixed (%)=0%, Features Unknown (%)=25%, Victim Classifier=Random Forest2025.0291
FRAUD-RLAFeatures Fixed (%)=25%, Features Unknown (%)=0%, FRAUD-RLA Time (Number of frauds)=1000, Victim Classifier=Random Forest2025.0291
FRAUD-RLAFeatures Fixed (%)=25%, Features Unknown (%)=0%, FRAUD-RLA Time (Number of frauds)=4000, Victim Classifier=Random Forest2025.0291
FRAUD-RLAFeatures Fixed (%)=0%, Features Unknown (%)=0%, FRAUD-RLA Time (Number of frauds)=1000, Victim Classifier=Random Forest2025.0290
FRAUD-RLAFeatures Fixed (%)=0%, Features Unknown (%)=25%, FRAUD-RLA Time (Number of frauds)=4000, Victim Classifier=Random Forest2025.0290
FRAUD-RLAFeatures Fixed (%)=25%, Features Unknown (%)=25%, FRAUD-RLA Time (Number of frauds)=4000, Victim Classifier=Random Forest2025.0290
FRAUD-RLAFeatures Fixed (%)=50%, Features Unknown (%)=0%, FRAUD-RLA Time (Number of frauds)=4000, Victim Classifier=Random Forest2025.0290
FRAUD-RLAFeatures Fixed (%)=0%, Features Unknown (%)=25%, FRAUD-RLA Time (Number of frauds)=1000, Victim Classifier=Random Forest2025.0289
FRAUD-RLAFeatures Fixed (%)=0%, Features Unknown (%)=50%, FRAUD-RLA Time (Number of frauds)=4000, Victim Classifier=Random Forest2025.0289
FRAUD-RLAFeatures Fixed (%)=25%, Features Unknown (%)=25%, FRAUD-RLA Time (Number of frauds)=1000, Victim Classifier=Random Forest2025.0288
FRAUD-RLAFeatures Fixed (%)=0%, Features Unknown (%)=0%, FRAUD-RLA Time (Number of frauds)=300, Victim Classifier=Random Forest2025.0287
FRAUD-RLAFeatures Fixed (%)=50%, Features Unknown (%)=0%, FRAUD-RLA Time (Number of frauds)=1000, Victim Classifier=Random Forest2025.0287
FRAUD-RLAFeatures Fixed (%)=25%, Features Unknown (%)=0%, FRAUD-RLA Time (Number of frauds)=300, Victim Classifier=Random Forest2025.0286
FRAUD-RLAFeatures Fixed (%)=0%, Features Unknown (%)=50%, FRAUD-RLA Time (Number of frauds)=1000, Victim Classifier=Random Forest2025.0286
FRAUD-RLAFeatures Fixed (%)=0%, Features Unknown (%)=25%, FRAUD-RLA Time (Number of frauds)=300, Victim Classifier=Random Forest2025.0283
FRAUD-RLAFeatures Fixed (%)=25%, Features Unknown (%)=25%, FRAUD-RLA Time (Number of frauds)=300, Victim Classifier=Random Forest2025.0277
FRAUD-RLAFeatures Fixed (%)=0%, Features Unknown (%)=50%, FRAUD-RLA Time (Number of frauds)=300, Victim Classifier=Random Forest2025.0276
FRAUD-RLAFeatures Fixed (%)=50%, Features Unknown (%)=0%, FRAUD-RLA Time (Number of frauds)=300, Victim Classifier=Random Forest2025.0276
Best BaselineFeatures Fixed (%)=0%, Features Unknown (%)=50%, Victim Classifier=Random Forest2025.0271
Best BaselineFeatures Fixed (%)=25%, Features Unknown (%)=25%, Victim Classifier=Random Forest2025.0271
FRAUD-RLAFeatures Fixed (%)=50%, Features Unknown (%)=25%, FRAUD-RLA Time (Number of frauds)=4000, Victim Classifier=Random Forest2025.0269
Best BaselineFeatures Fixed (%)=50%, Features Unknown (%)=0%, Victim Classifier=Random Forest2025.0265
FRAUD-RLAFeatures Fixed (%)=25%, Features Unknown (%)=50%, FRAUD-RLA Time (Number of frauds)=4000, Victim Classifier=Random Forest2025.0264
FRAUD-RLAFeatures Fixed (%)=50%, Features Unknown (%)=25%, FRAUD-RLA Time (Number of frauds)=1000, Victim Classifier=Random Forest2025.0262
FRAUD-RLAFeatures Fixed (%)=25%, Features Unknown (%)=50%, FRAUD-RLA Time (Number of frauds)=1000, Victim Classifier=Random Forest2025.0257
FRAUD-RLAFeatures Fixed (%)=50%, Features Unknown (%)=25%, FRAUD-RLA Time (Number of frauds)=300, Victim Classifier=Random Forest2025.0248
FRAUD-RLAFeatures Fixed (%)=25%, Features Unknown (%)=50%, FRAUD-RLA Time (Number of frauds)=300, Victim Classifier=Random Forest2025.0246
Best BaselineFeatures Fixed (%)=50%, Features Unknown (%)=25%, Victim Classifier=Random Forest2025.0235
Best BaselineFeatures Fixed (%)=25%, Features Unknown (%)=50%, Victim Classifier=Random Forest2025.0234