There are numerous applications to Bayesian machine learning methods to investment. The transparency that probabilistic estimates create are naturally useful for risk management and performance evaluation. We will illustrate the computation and comparison of a metric like the Sharpe ratio. The GitHub repository also includes two notebooks referenced below that present the use of Bayesian ML for modeling linear time series and stochastic volatility.

These notebooks have been adapted from tutorials created at Quantopian where Thomas Wiecki leads data science and has significantly contributed to popularizing the use of Bayesian methods. The references also include a tutorial on using Bayesian ML to estimate pairs trading hedging ratios.