Usage.txt

# Fit and predict with logistic regression
log_reg = LogisticRegression(random_state=42)
log_reg.fit(X_train_scaled, y_train)
y_pred = log_reg.predict(X_test_scaled)

# Calculate and plot confusion matrix
conf_matrix = confusion_matrix(y_test, y_pred)
plt.figure(figsize=(8, 4))
sns.heatmap(conf_matrix, annot=True, fmt='d', cmap='Blues', xticklabels=['No AR', 'AR'], yticklabels=['No AR', 'AR'])
plt.xlabel('Predicted')
plt.ylabel('Actual')
plt.title('Confusion Matrix')
plt.show()

# Calculate total cost with optimal threshold
optimal_cost, cm = calculate_costs(y_test, best_model_proba, optimal_threshold)

# Compare costs
labels = ['With Model', 'Without Model']
costs = [optimal_cost, labor_cost_without_model]
plt.bar(labels, costs, color=['skyblue', 'orange'])
plt.ylabel('Total Cost (EUR)')
plt.title('Business Impact')
plt.show()

# Cost vs. Threshold Plot
costs = []
thresholds = np.arange(0.0, 1.0, 0.01)
for threshold in thresholds:
    total_cost, _ = calculate_costs(y_test, best_model_proba, threshold)
    costs.append(total_cost)

plt.figure(figsize=(8, 6))
plt.plot(thresholds, costs, color='blue', label='Total Cost')
plt.axvline(x=optimal_threshold, color='red', linestyle='--', label=f'Optimal Threshold ({optimal_threshold})')
plt.xlabel('Threshold')
plt.ylabel('Total Cost (EUR)')
plt.title('Total Cost vs. Threshold')
plt.legend(loc='upper right')
plt.show()