Back to Projects
Complete
Featured
LSTM Stock Price Prediction
Python
TensorFlow
Keras
LSTM
Neural Networks
Machine Learning
Time Series
About this project
A neural network-based stock price prediction system using Long Short-Term Memory (LSTM) networks. Built as a final project for CS 682 Neural Networks course at UMass Amherst. Implements a multi-layer LSTM architecture with dropout regularization, layer normalization, and early stopping for predicting daily stock closing prices.
Highlights
- Multi-layer LSTM architecture with 4 LSTM layers and dropout regularization
- Custom PriceClassifier class for stock price prediction and evaluation
- Train/validation/test split with temporal data handling
- Early stopping callback to prevent overfitting
- Visualization of predicted vs actual stock prices
- Mean squared error evaluation metrics
- Support for layer normalization and configurable hyperparameters
- Jupyter notebook implementation with comprehensive analysis