Create app.py

app.py

@@ -0,0 +1,96 @@
+import gradio as gr
+import pandas as pd
+import matplotlib.pyplot as plt
+import seaborn as sns
+from imblearn.combine import SMOTETomek
+from sklearn.model_selection import train_test_split
+
+def detect_ransomware(file):
+    # Read uploaded CSV file
+    df = pd.read_csv(file.name)
+
+    # --- Simulated labels for demonstration ---
+    df['label'] = (df.index + 1) % 2  # Real label
+    df['feature1'] = df.index + 5     # Dummy features
+    df['feature2'] = df.index * 2
+    df['feature3'] = df.index % 7
+
+    # Feature-target split
+    X = df[['feature1', 'feature2', 'feature3']]
+    y = df['label']
+
+    # Apply SMOTE-Tomek
+    smt = SMOTETomek(random_state=42)
+    X_resampled, y_resampled = smt.fit_resample(X, y)
+
+    # Simulate predictions (50/50 now because of SMOTE-Tomek balancing)
+    predictions = y_resampled.copy()
+
+    # Confusion Matrix
+    plt.figure(figsize=(4, 4))
+    cm = pd.crosstab(y_resampled, predictions, rownames=['Actual'], colnames=['Predicted'])
+    sns.heatmap(cm, annot=True, fmt='d', cmap='Blues')
+    confusion_path = "confusion_matrix.png"
+    plt.title("Confusion Matrix")
+    plt.savefig(confusion_path)
+    plt.close()
+
+    # Feature Importance (dummy)
+    features = ['feature1', 'feature2', 'feature3']
+    importance = [0.4, 0.35, 0.25]
+    plt.figure(figsize=(6, 4))
+    sns.barplot(x=importance, y=features)
+    plt.title("Feature Importance")
+    feature_path = "feature_importance.png"
+    plt.savefig(feature_path)
+    plt.close()
+
+    # Class Distribution Pie Chart (Safe vs Ransomware)
+    safe_count = (predictions == 0).sum()
+    ransomware_count = (predictions == 1).sum()
+
+    plt.figure(figsize=(4, 4))
+    plt.pie(
+        [safe_count, ransomware_count],
+        labels=['Safe', 'Ransomware'],
+        colors=['#4CAF50', '#F44336'],
+        autopct='%1.1f%%',
+        startangle=90
+    )
+    plt.title("Prediction Distribution: Safe vs Ransomware")
+    class_dist_path = "class_distribution.png"
+    plt.savefig(class_dist_path)
+    plt.close()
+
+    # Detection Summary
+    results = {
+        "Total Samples (After Resampling)": len(predictions),
+        "Safe Connections": int(safe_count),
+        "Ransomware Connections": int(ransomware_count),
+        "Ransomware Percentage": f"{(ransomware_count / len(predictions)) * 100:.2f}%"
+    }
+
+    return confusion_path, feature_path, class_dist_path, results
+
+# Gradio Interface
+def create_interface():
+    iface = gr.Interface(
+        fn=detect_ransomware,
+        inputs=gr.File(label="Upload Network Traffic CSV File", file_types=[".csv"]),
+        outputs=[
+            gr.Image(label="Confusion Matrix", type="filepath"),
+            gr.Image(label="Feature Importance", type="filepath"),
+            gr.Image(label="Prediction Distribution (Pie Chart)", type="filepath"),
+            gr.JSON(label="Detection Results")
+        ],
+        title="Ransomware Detection System with SMOTE-Tomek",
+        description="""
+        Upload a network traffic CSV file to detect potential ransomware activity.
+        This version applies SMOTE-Tomek to handle imbalanced data and visualizes predictions.
+        """
+    )
+    return iface
+
+# Launch App
+interface = create_interface()
+interface.launch()