Deep Learning for Finance

Validation Design for Deep Learning Trading Strategies

Study how chronological splits, walk-forward validation, and label purging change apparent performance for MLP, LSTM, and CNN-GAF models on Bitcoin returns.