An interactive machine learning web application that predicts house prices using neural networks. Built with TensorFlow.js for real-time AI predictions in the browser.
This project demonstrates a complete machine learning pipeline that:
- Trains a neural network to predict house prices
- Uses real-time data visualization during training
- Provides instant AI-powered price predictions
- Shows model performance and accuracy metrics
- 🧠 Neural Network Training - Watch AI learn in real-time
- 📊 Live Predictions - Get instant house price estimates
- 📈 Training Visualization - See loss reduction over epochs
- 🎨 Modern UI - Clean, responsive design
- 📱 Mobile Friendly - Works on all devices
- ⚡ Fast Performance - Browser-based ML with no server needed
- TensorFlow.js - Machine learning framework
- Chart.js - Data visualization
- JavaScript ES6 - Core programming
- HTML5 & CSS3 - Frontend development
- Neural Networks - Deep learning model
- Creates 500+ synthetic house records
- Features: bedrooms, bathrooms, square footage, age, location
- Realistic price calculations with market factors
- Data normalization and scaling
- 30 epochs with batch training
- Real-time loss tracking
- Progressive model improvement
- Takes user input for house features
- Normalizes data using training parameters
- Runs inference through trained model
- Shows price estimate with confidence score
- Clone the repository
git clone https://github.com/yourusername/ai-house-price-predictor.git cd ai-house-price-predictor
ai-house-price-predictor/
│
├── index.html # Main application html file
├── style.css # css file
└── script.js # java script file