ritesh1918/house-price-predictor

Linear Regression House Price Predictor

A simple LinearRegression model trained on a synthetic housing dataset. Built as a teaching artifact to demonstrate an end-to-end Machine Learning engineering workflow (data generation -> model training -> API serving -> frontend UI).

Model Details

• Algorithm: Ordinary Least Squares (OLS) Linear Regression (scikit-learn)
• Features used:
  • size_sqft: Size of the property in square feet (float)
  • bedrooms: Number of bedrooms (int)
  • bathrooms: Number of bathrooms (int)
  • age_years: Age of the property in years (int)
  • location_score: Neighborhood desirability score from 1-10 (int)
• Target: price_lakhs: Price of the property in Lakhs (float)

Training Data & Parameters

The training dataset is synthetically generated (300 samples) with features linearly combined plus Gaussian noise (mean=0, std=12) to simulate realistic market variability.

• Train/Test Split: 80/20
• Random State: 42

Performance Metrics

• Mean Absolute Error (MAE): 8.78 Lakhs
• Root Mean Squared Error (RMSE): 11.23 Lakhs
• R² Score: 0.9566

How to Use

import pickle
import numpy as np

# Load model
with open("model.pkl", "rb") as f:
    model = pickle.load(f)

# Input format: [size_sqft, bedrooms, bathrooms, age_years, location_score]
sample_input = np.array([[1500, 3, 2, 5, 8]])
prediction = model.predict(sample_input)[0]
print(f"Predicted Price: {prediction:.2f} Lakhs")