Machine learning is here and being used by a broad array of firms. This is the strong conclusion from a recent survey report by the Bank of England and the FCA called "Machine Learning in UK Financial Services". Two thirds of the respondents have machine learning tool in use. This is not just the study of the processes but actual usage across a wide set of problems. Firms must like the results because there is a strong development pipeline for more machine learning applications.
The places where the greatest number machine learning applications are in full deployment include risk management, compliance and customer engagement.
However, there are constraints with using machine learning applications. It is not a regulatory issue but legacy systems that may not be able to handle the computing and data issues.
All of these firms follow similar processes with how models are developed.
The set of techniques that have been classified as machine learning is vast; however, it seems that tree-based models that can handle non-linear relationships are the front-runner for applications.
If you are not at least looking at how to incorporate machine learning in your business processes, you are falling behind your competition.
No comments:
Post a Comment