Skip to main content
ARCHIVED
  • Python
  • PyTorch
  • CUDA
  • yfinance
  • pytest
  • uv
STARTED NOV 11, 2024

Stock Price Prediction using LSTM, RNN, and CNN-Sliding-Window Models

A PyTorch reproduction of the paper "Stock Price Prediction using LSTM, RNN and CNN-Sliding Window Model" (Hiransha et al.), built as a Master's degree final project for the course Intelligent Algorithms.

Stock Price Prediction with Deep Learning

A PyTorch reproduction of the IEEE paper Stock Price Prediction using LSTM, RNN and CNN-Sliding Window Model, built as my final project for Intelligent Algorithms during my Masters at Johns Hopkins.

Benchmarks a vanilla RNN, an LSTM, and a 1D-CNN over sliding windows against an ARIMA baseline, so the neural models have to beat a classical one rather than just each other. Structured as an installable package with configuration-driven experiments, a shared training loop, and a pytest suite.

Not a strict replication — the paper uses multi-year daily closes, this runs on five days of intraday data so experiments finish fast.

This project is not actively hosted anywhere therefore it's archived.