Andrew Li Discusses Beyond the Black Box - Machine Learning for Equities

December 14, 2022

Developing reliable and intuitive interpretation is essential for the application of machine learning to investing. This presentation discussed a framework for decomposing any machine learning model into linear, nonlinear, and interaction effects that drive both prediction and performance. With a case study of predicting US large cap stock returns, this presentation showed how the "Model Fingerprint" tool enables practitioners to summarize key characteristics, similarities and differences among different models, thereby enhancing their understanding of the market.