Cornell-Citi Financial Data Science Webinar with Michael Rabadi

to

Location

Description

The Cornell-Citi Financial Data Science Webinars will commence in the fall semester with Michael Rabadi (Balyasny Asset Management). He will be giving a talk titled "Online Learning and Model Selection for Financial Time Series" on Tuesday, 9/1. This webinar be held via Zoom from 5:00pm to 6:00pm. Seminars are free. However, in order to receive the Zoom link, registration is required, and this event is restricted to the Cornell community (sign up with Cornell email): https://cornell.zoom.us/webinar/register/WN_9dJSu1PxQJytrcx7oEBIQw Abstract: Model selection is one of the most important problems in machine learning. While common methods, like cross validation, can be applied to many domains, it is usually unsuitable when selecting a model for a financial time series. During this seminar we will cover common model selection algorithms and evaluate their assumptions. We will then show why these algorithms are typically unsuited for financial time series. We will then show how online learning can be used to address the problem at hand. We will conclude with sample code and an example using real Bitcoin data. Bio: Michael Rabadi is the Lead Machine Learning Researcher at Balyasny Asset Management, where he leads machine learning efforts across the firm. Prior to BAM, he was the Head of Machine Learning and a Portfolio Manager at quantPORT, where he led machine learning efforts across the quantitative investment process. Prior to quantPORT, he held various technical, research, and managerial roles, including ones with Spotify, NYU, and various startups.