AI House Prediction Web App

Models

The following models were used to predict californian house prices and evaluated:

• •
  ✅XGBoost
• •
  Neural Network
• •
  Linear Regression
• •
  Random Forest Regressor

🏆 XGBOOST achieved lowest rmse: 47989.31 (root mean squared error) so was chosen for website.

Fast API

Basic FastAPI server is used to serve the XGBoost model and predict house prices.

Endpoint

POST /predict

Sample Request Body:

{
  "longitude": -122.23,
  "latitude": 37.88,
  "housing_median_age": 41,
  "total_rooms": 880,
  "total_bedrooms": 129,
  "population": 322,
  "households": 126,
  "median_income": 8.3252,
  "ocean_proximity": "NEAR BAY"
}

Sample Response:

{
  "model": "xgboost",
  "predicted_value": 416807.44
}

Streamlit Web App

The front-end is built using Streamlit, providing a simple UI for users to input housing features and get real-time predictions.

Features:

• •
  User input form for housing characteristics
• •
  Location preview using an interactive map
• •
  Integration with FastAPI to fetch prediction results
• •
  Responsive and clean UI

Available here.