Personal Website
stock_hmm.py downloads historical prices for the top 20 S&P 500 stocks
and trains a Gaussian Hidden Markov Model on the log returns for each
ticker. The script prints the expected next-day return as a simple
directional indicator.
The script depends on numpy, pandas, yfinance, hmmlearn, and
scikit-learn.
python stock_hmm.py --start 2018-01-01Optional --tickers arguments can be provided to override the default
top 20 list.