Deep learning project for stock market prediction using historical price data. Compares single-stock and multi-stock time-series models (e.g., LSTM/Transformer) to test whether incorporating data from multiple correlated stocks improves prediction accuracy. Includes preprocessing, training, and evaluation pipelines in PyTorch.