app.py

import gradio as gr
import pandas as pd
import numpy as np
import joblib
from sklearn.ensemble import RandomForestClassifier
from sklearn.preprocessing import StandardScaler

# Load the dataset for feature extraction
csv_file_path = "creditcard.csv"  # Ensure this file is uploaded to Hugging Face
df = pd.read_csv(csv_file_path)

# Select important features
selected_features = ["Amount", "V1", "V2", "V3", "Class"]
df = df[selected_features]

# Train-test split
X = df.drop(columns=["Class"])
y = df["Class"]

# Standardize "Amount"
scaler = StandardScaler()
X["Amount"] = scaler.fit_transform(X[["Amount"]])

# Train model
model = RandomForestClassifier(n_estimators=100, random_state=42)
model.fit(X, y)

# Save model and scaler
joblib.dump(model, "fraud_model.pkl")
joblib.dump(scaler, "scaler.pkl")


# Define function for prediction
def predict_fraud(amount, v1, v2, v3):
    scaler = joblib.load("scaler.pkl")
    model = joblib.load("fraud_model.pkl")

    input_data = np.array([[amount, v1, v2, v3]])
    input_data[:, 0] = scaler.transform(input_data[:, [0]])  # Scale amount

    prediction = model.predict(input_data)
    return "Fraudulent Transaction" if prediction[0] == 1 else "Non-Fraudulent Transaction"


# Gradio UI
iface = gr.Interface(
    fn=predict_fraud,
    inputs=[
        gr.Number(label="Transaction Amount"),
        gr.Number(label="V1"),
        gr.Number(label="V2"),
        gr.Number(label="V3"),
    ],
    outputs="text",
    title="Credit Card Fraud Detection",
    description="Enter transaction details to check if it's fraudulent or not.",
)

if __name__ == "__main__":
    iface.launch()