The Big Challenge in Machine Learning:

Machine Learning (ML) is one of the hottest buzz words these days and has some really fancy applications across industries. However, there is a significant impedance in the acceptance of ML models in regulated industries, mainly financial services and pharma. This is primarily driven by the lack of adequate explainability of ML models, both from the top leadership and the regulatory perspective.

BlackRock (the world’s largest asset manager with ~6.5 trillion dollars asset under management) spent multi-million dollars to develop really powerful neural-network-based liquidity models that significantly outperformed the traditional liquidity models. The firm, however, had to shelve those as the developers weren’t able to satisfactorily explain their working to the top leadership.

ML explainability is an area of research that is gaining a lot of traction. The key objective herein is to make the ML models more transparent, fair, ethical, and accountable, and thus more acceptable.

For those interested in exploring this high-demand area, I’ve attached a paper that I read a year back. Thought to share as it provides a nice overview of ML explainability and is good enough to give a jump-start on the topic.

Keywords : #Machine Learning Institute, #Machine Learning Certification, #Machine Learning Courses #Machine Learning Training #Machine Learning in Finance #Machine Learning in Quantitative Finance #Machine Learning Overview #Machine Learned #Machine Classes

#machinelearning #financialservices

Our Machine Learning Facebook Page :

Tags: ,