Summary
I enjoyed discussing with Igor the uptake and increasing adoption of ML in finance and, in particular, the application of reinforcement learning. I think it’s worthwhile for all data scientists who work in the broader financial services field to investigate opportunities for the application of reinforcement learning. It was also great to discuss the importance of explainable AI in finance, and the challenges involved.
Igor talked about the value of seeing the world and specific problems from different perspectives. This principle operates on a number of levels – it can be achieved by building teams that include people with backgrounds in different disciplines, but it can also mean recreating a model using a different class of modelling – for instance applying a ML model to a problem generally represented by a linear model.
Regarding high failure rates of AI projects, Igor’s view is that as most ideas don’t end up being successful (as...