Fraud Detection on Dataset 1
100PerformanceBest Baseline
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Best BaselineFixed Features (%)=0 %, Unknown Features (%)=0 %, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | 100 | — | |
| SKLearn Neural Network (Baseline)Model Architecture=Neural Network, Fixed Features Percentage=0%, Unknown Features Percentage=0%2025.02 | 100 | — | |
| Best BaselineFixed Features (%)=25%, Unknown Features (%)=0%, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | 99 | — | |
| Best BaselineFixed Features (%)=0 %, Unknown Features (%)=25%, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | 97 | — | |
| SKLearn Neural Network (Baseline)Model Architecture=Neural Network, Fixed Features Percentage=25%, Unknown Features Percentage=0%2025.02 | 97 | — | |
| SKLearn Neural Network (Baseline)Model Architecture=Neural Network, Fixed Features Percentage=0%, Unknown Features Percentage=25%2025.02 | 91 | — | |
| Best BaselineFixed Features (%)=50%, Unknown Features (%)=0%, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | 58 | — | |
| Best BaselineFixed Features (%)=0 %, Unknown Features (%)=50%, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | 53 | — | |
| SKLearn Neural Network (Baseline)Model Architecture=Neural Network, Fixed Features Percentage=50%, Unknown Features Percentage=0%2025.02 | 45 | — | |
| Best BaselineFixed Features (%)=25%, Unknown Features (%)=25%, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | 43 | — | |
| SKLearn Neural Network (Baseline)Model Architecture=Neural Network, Fixed Features Percentage=25%, Unknown Features Percentage=25%2025.02 | 42 | — | |
| SKLearn Neural Network (Baseline)Model Architecture=Neural Network, Fixed Features Percentage=0%, Unknown Features Percentage=50%2025.02 | 39 | — | |
| SKLearn Neural Network (Baseline)Model Architecture=Neural Network, Fixed Features Percentage=50%, Unknown Features Percentage=25%2025.02 | 4 | — | |
| Best BaselineFixed Features (%)=25%, Unknown Features (%)=50%, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | 3 | — | |
| SKLearn Neural Network (Baseline)Model Architecture=Neural Network, Fixed Features Percentage=25%, Unknown Features Percentage=50%2025.02 | 2 | — | |
| Best BaselineFixed Features (%)=50%, Unknown Features (%)=25%, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | 1 | — | |
| FRAUD-RLAFixed Features (%)=0 %, Unknown Features (%)=0 %, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | — | -0.12 | |
| FRAUD-RLAFixed Features (%)=0 %, Unknown Features (%)=25%, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | — | -0.12 | |
| FRAUD-RLAFixed Features (%)=25%, Unknown Features (%)=0%, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | — | -0.11 | |
| FRAUD-RLAFixed Features (%)=0 %, Unknown Features (%)=50%, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | — | 0.08 | |
| FRAUD-RLAFixed Features (%)=25%, Unknown Features (%)=25%, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | — | 0.17 | |
| FRAUD-RLAFixed Features (%)=50%, Unknown Features (%)=0%, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | — | 0.05 | |
| FRAUD-RLAFixed Features (%)=25%, Unknown Features (%)=50%, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | — | 0 | |
| FRAUD-RLAFixed Features (%)=50%, Unknown Features (%)=25%, Classifier=Random Forest, Measurement Window=4000 frauds2025.02 | — | 0 | |
| FRAUD-RLAModel Architecture=Neural Network, Fixed Features Percentage=0%, Unknown Features Percentage=0%2025.02 | — | -0.08 | |
| FRAUD-RLAModel Architecture=Neural Network, Fixed Features Percentage=0%, Unknown Features Percentage=25%2025.02 | — | 0.02 | |
| FRAUD-RLAModel Architecture=Neural Network, Fixed Features Percentage=25%, Unknown Features Percentage=0%2025.02 | — | 0.04 | |
| FRAUD-RLAModel Architecture=Neural Network, Fixed Features Percentage=0%, Unknown Features Percentage=50%2025.02 | — | 0.54 | |
| FRAUD-RLAModel Architecture=Neural Network, Fixed Features Percentage=25%, Unknown Features Percentage=25%2025.02 | — | 0.51 | |
| FRAUD-RLAModel Architecture=Neural Network, Fixed Features Percentage=50%, Unknown Features Percentage=0%2025.02 | — | 0.49 | |
| FRAUD-RLAModel Architecture=Neural Network, Fixed Features Percentage=25%, Unknown Features Percentage=50%2025.02 | — | 0.94 | |
| FRAUD-RLAModel Architecture=Neural Network, Fixed Features Percentage=50%, Unknown Features Percentage=25%2025.02 | — | 0.92 |