Instructions to use jorgeasmz/fraud-stream-detector with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Scikit-learn
How to use jorgeasmz/fraud-stream-detector with Scikit-learn:
from huggingface_hub import hf_hub_download import joblib model = joblib.load( hf_hub_download("jorgeasmz/fraud-stream-detector", "sklearn_model.joblib") ) # only load pickle files from sources you trust # read more about it here https://skops.readthedocs.io/en/stable/persistence.html - Notebooks
- Google Colab
- Kaggle
| { | |
| "threshold": 0.01643005512716834, | |
| "daily_budget": 100, | |
| "training_days": 91, | |
| "training_rows": 872795, | |
| "held_out_rows": 813843, | |
| "held_out_days": 85, | |
| "alerts_per_day_at_threshold": 120.1, | |
| "precision_at_threshold": 0.5319, | |
| "card_precision_at_budget": 0.5613, | |
| "precision_at_budget": 0.6195, | |
| "scenario_recall": { | |
| "1": 0.9161, | |
| "2": 0.6643, | |
| "3": 0.8191 | |
| }, | |
| "columns": [ | |
| "amount", | |
| "hour", | |
| "is_weekend", | |
| "customer_count_1d", | |
| "customer_mean_1d", | |
| "customer_count_7d", | |
| "customer_mean_7d", | |
| "customer_count_30d", | |
| "customer_mean_30d", | |
| "terminal_count_1d", | |
| "terminal_mean_1d", | |
| "terminal_count_7d", | |
| "terminal_mean_7d", | |
| "terminal_count_30d", | |
| "terminal_mean_30d", | |
| "amount_over_customer_mean_1d", | |
| "amount_over_customer_mean_7d", | |
| "amount_over_customer_mean_30d", | |
| "customer_risk_1d", | |
| "customer_risk_count_1d", | |
| "customer_risk_7d", | |
| "customer_risk_count_7d", | |
| "customer_risk_30d", | |
| "customer_risk_count_30d", | |
| "terminal_risk_1d", | |
| "terminal_risk_count_1d", | |
| "terminal_risk_7d", | |
| "terminal_risk_count_7d", | |
| "terminal_risk_30d", | |
| "terminal_risk_count_30d" | |
| ] | |
| } | |