import pandas as pd import gradio as gr from sklearn.model_selection import train_test_split from sklearn.ensemble import RandomForestClassifier from sklearn.preprocessing import LabelEncoder # Load dataset df = pd.read_csv("fraud_data.csv") # Preprocess data le_country = LabelEncoder() le_device = LabelEncoder() df['country'] = le_country.fit_transform(df['country']) df['device'] = le_device.fit_transform(df['device']) # Convert transaction_time to minutes df['transaction_time'] = df['transaction_time'].apply(lambda x: int(x.split(":")[0]) * 60 + int(x.split(":")[1])) # Features and target X = df.drop('is_fraud', axis=1) y = df['is_fraud'] # Split data X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42) # Train model model = RandomForestClassifier() model.fit(X_train, y_train) # Prediction function def predict_fraud(amount, country, device, transaction_time): country_encoded = le_country.transform([country])[0] device_encoded = le_device.transform([device])[0] time_in_minutes = int(transaction_time.split(":")[0]) * 60 + int(transaction_time.split(":")[1]) input_data = [[amount, country_encoded, device_encoded, time_in_minutes]] prediction = model.predict(input_data)[0] return "Fraudulent" if prediction == 1 else "Legitimate" # Gradio interface countries = df['country'].unique() devices = df['device'].unique() iface = gr.Interface( fn=predict_fraud, inputs=[ gr.Number(label="Amount"), gr.Dropdown(choices=le_country.classes_.tolist(), label="Country"), gr.Dropdown(choices=le_device.classes_.tolist(), label="Device"), gr.Textbox(label="Transaction Time (HH:MM)") ], outputs="text", title="Fraud Detection System", description="Enter transaction details to predict if it's fraudulent." ) iface.launch()