A Review of Risks Associated with Machine Learning in Application to Quantitative Investment

Abstract

Based on inspirations and ideas from relevant literatures, this paper evaluated the risks associated with using random forest, XG Boost and logistic regression for quantitative investment from the perspective of its accuracy, adaptability, efficiency, simplicity and interpretability. Overall, the random forest and the XG Boost contains better accuracy and have higher adaptability than the logistic regression as they are susceptible to different data types. The XG Boost have the fastest processing speed which gives it higher efficiency over the other two, however it is also the most difficult to implement as it is written in C++. All three algorithms are relatively clear and easy to understand. This work hopes to assist investors in their decision making on which model to use.