Model Training

# Define classifiers 
classifiers = [
    LogisticRegression(max_iter=1000),
    GradientBoostingClassifier(),
    SVC(kernel='linear', probability=True),
    SVC(kernel='rbf', probability=True),
    KNeighborsClassifier(),
    DecisionTreeClassifier(),
    RandomForestClassifier(),
    MLPClassifier(max_iter=1000)
]

# Classifier names
classifier_names = [
    "Logistic Regression", "Gradient Boosting", "SVM (Linear)", "SVM (RBF)",
    "K-Nearest Neighbours (k-NN)", "Decision Tree", "Random Forest", "Neural Network"
]

# Initialize storage for metrics
results = []

# Fit, evaluate, and visualize classifiers 
for clf, name in zip(classifiers, classifier_names):
    clf.fit(X_train, y_train)
    y_pred = clf.predict(X_test)
    y_prob = clf.predict_proba(X_test)[:, 1] if hasattr(clf, "predict_proba") else None

    # Collect metrics
    accuracy = accuracy_score(y_test, y_pred)
    precision = precision_score(y_test, y_pred)
    recall = recall_score(y_test, y_pred)
    f1 = f1_score(y_test, y_pred)
    roc_auc = roc_auc_score(y_test, y_prob) if y_prob is not None else None

    results.append({
        "Model": name,
        "Accuracy": round(accuracy,4),
        "Precision_Score": round(precision,4),
        "Recall_Score": round(recall,4),
        "F1_Score": round(f1,4),
        "ROC_AUC": round(roc_auc,4)
    })

# Create DataFrame
comparison_table = pd.DataFrame(results)

# Print the comparison table
print(comparison_table)

# Plot the table
fig, ax = plt.subplots(figsize=(14, 8))  # Set the figure size
ax.axis('tight')
ax.axis('off')

# Table
table = ax.table(cellText=comparison_table.values, colLabels=comparison_table.columns, cellLoc='center', loc='center')

# Styling the table
table.auto_set_font_size(False)
table.set_fontsize(14)
table.scale(1.5, 1.5)

# Set colors for header
for j in range(len(comparison_table.columns)):
    table[0, j].set_fontsize(12)
    table[0, j].set_text_props(weight='bold')
    table[0, j].set_facecolor('#40466e')
    table[0, j].set_text_props(color='white')

# Define the indices of the rows you want to highlight 
highlight_rows = [1, 3, 5, 7]  # Change these to the indices of the rows you want to highlight

for highlight_row in highlight_rows:
    for i in range(len(comparison_table.columns)):
        table[highlight_row, i].set_facecolor('#c5c5e5')

plt.show()