Simple Flask web app to deploy a model

Key Concepts

• Train and save machine learning models (pickle/joblib)
• Load the saved model inside Flask to avoid reloading on each request
• Flask routes handle requests and responses
• Input validation and JSON-based prediction responses
• Extendable for production use with Gunicorn, Docker, etc.

1. Train and Save Model (Python Script)

from sklearn.datasets import load_iris
from sklearn.ensemble import RandomForestClassifier
import pickle

# Load example dataset
data = load_iris()
X, y = data.data, data.target

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

# Save the trained model to disk
with open('model.pkl', 'wb') as f:
    pickle.dump(model, f)

print("Model saved successfully.")

2. Flask App to Load & Serve the Saved Model

from flask import Flask, request, jsonify
import pickle
import numpy as np

app = Flask(__name__)

# Load the saved model once when app starts
with open('model.pkl', 'rb') as f:
    model = pickle.load(f)

@app.route('/')
def home():
    return "Welcome to the Iris Classifier API!"

@app.route('/predict', methods=['POST'])
def predict():
    try:
        # Parse JSON input
        data = request.get_json(force=True)
        features = data['features']

        # Convert to numpy array and reshape for prediction
        input_data = np.array(features).reshape(1, -1)

        # Predict class
        prediction = model.predict(input_data)

        # Return predicted class label and probabilities
        prediction_proba = model.predict_proba(input_data)
        return jsonify({
            'prediction': int(prediction[0]),
            'probabilities': prediction_proba[0].tolist()
        })

    except Exception as e:
        return jsonify({'error': str(e)})

if __name__ == '__main__':
    app.run(debug=True)