Machine Factor Technologies

Great job done by our CIO, Александр Проскурин! His paper "Does the CFTC Report Have Predictive Power: Machine Learning Approach" was accepted to be added into summer issue of Journal of Financial Data Science (Portfolio Management Research).
The paper applies financial machine learning techniques to investigate if the COT report information can be used to predict the prices of most actively traded agricultural futures.
https://jfds.pm-research.com/content/early/2021/06/10/jfds.2021.1.065

On June 4, our Founder and CIO, Александр Проскурин (Alexandr Proskurin) will be speaking at Big Data Finance 2020 conference on optimal trading rules detection. As a part of the lecture, Oleksandr will show the example of applying trading rules detection under capital constraints for VIX futures trading strategy.

Interpreting the results of machine learning model is a key to successful strategy research. There are various feature importance algorithms such as MDI, MDA and SFI. Recently, our research team helped Hudson & Thames Quantitative Research to enrich package feature importance module with Model Fingerprints algorithm

https://hudsonthames.org/interpreting-machine-learning-model-fingerprints-algorithm/