fraud_detection / plot_metrics.py
VSaiTejaReddy004's picture
feat: clean initial commit with git lfs
07ff411
Raw History Blame Contribute Delete
2.41 kB
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay, roc_curve, auc
import joblib
import os
CSV_PATH = "creditcard.csv"
MODEL_PATH = "model.pkl"
def generate_graphs():
if not os.path.exists(MODEL_PATH) or not os.path.exists(CSV_PATH):
print(f"Missing {MODEL_PATH} or {CSV_PATH}. Please run train_model.py first.")
return
print("Loading Machine Learning model...")
clf = joblib.load(MODEL_PATH)
df = pd.read_csv(CSV_PATH)
from sklearn.model_selection import train_test_split
X = df.drop(columns=["Class", "Time"])
y = df["Class"]
_, X_test, _, y_test = train_test_split(X, y, test_size=0.2, random_state=42, stratify=y)
y_pred = clf.predict(X_test)
y_proba = clf.predict_proba(X_test)[:, 1]
importances = clf.feature_importances_
features = X.columns
idx = np.argsort(importances)[::-1]
sorted_features = [features[i] for i in idx]
sorted_importances = [importances[i] for i in idx]
cm = confusion_matrix(y_test, y_pred)
fpr, tpr, _ = roc_curve(y_test, y_proba)
roc_auc = auc(fpr, tpr)
fig, axes = plt.subplots(1, 3, figsize=(18, 6))
# 1. Feature Importance
axes[0].barh(sorted_features[:15][::-1], sorted_importances[:15][::-1], color='#4C72B0')
axes[0].set_title("Top 15 Feature Importances (What it looks at)")
axes[0].set_xlabel("Importance Score")
# 2. Confusion Matrix
disp = ConfusionMatrixDisplay(confusion_matrix=cm, display_labels=["Normal", "Fraud"])
disp.plot(ax=axes[1], cmap="Blues", colorbar=False)
axes[1].set_title("Confusion Matrix (Right vs Wrong)")
# 3. ROC Curve
axes[2].plot(fpr, tpr, color='darkorange', lw=2, label=f'ROC curve (AUC = {roc_auc:.2f})')
axes[2].plot([0, 1], [0, 1], color='navy', lw=2, linestyle='--')
axes[2].set_xlim([0.0, 1.0])
axes[2].set_ylim([0.0, 1.05])
axes[2].set_xlabel('False Positive Rate')
axes[2].set_ylabel('True Positive Rate')
axes[2].set_title('Receiver Oper. Characteristic (Precision curve)')
axes[2].legend(loc="lower right")
plt.tight_layout()
output_file = "model_evaluation_dashboard.png"
plt.savefig(output_file, dpi=150)
print(f"Successfully generated {output_file}!")
if __name__ == "__main__":
generate_graphs()