Speaker "Mikey Shulman" Details Back



Machine Learning vs Finance


Data Scientists and machine learning practitioners often find that the finance industry poses a unique set of challenges for analysis. Many challenges stem from the fact that an extremely wide breadth of data are relevant to finance professionals, including satellite imagery, news and asset price timeseries. Because money is on the line, and is often staked by humans (instead of algorithms), machine learning models must be highly accurate and interpretable. I will review some of these challenges and describe Kensho's approach to some practical, finance-specific solutions. With these technologies, we are able to structure the unstructured and help the finance professional make sense of the messy world.


Mikey heads the machine learning team at Kensho, which tackles problems across the finance and intelligence and defense industries. The machine learning team at Kensho specializes in a wide range of problems including novel approaches to timeseries prediction, behavioral economics, and industry-specific natural language processing. Prior to joining Kensho Mikey received a PhD in physics from Harvard University where he applied principles of machine learning to improve quantum computing and used natural language processing to identify, understand, and predict trends in physics research topics.