juanda-code269/FraudLens

Build TracePoint in Python/Streamlit: upload/demo fraud data, EDA, Logistic Regression/Random Forest/Gradient Boosting, PR-AUC/ROC-AUC, threshold tuning, false-positive/false-negative costs, optimal expected-loss threshold, single/batch scoring, feature explanations, tests, and clear limitations.

语言：Python · GitHub 星标：48 · Fork：28

在 GitHub 上查看原仓库