Perils of Quantitative Investing

IMPORTANT DETAIL: Our host is requiring attendee first and last names. If this information is not available from your meetup.com account, please enter it into this google doc (LINK (https://docs.google.com/forms/d/1ooKjM1nLucXGSSL4kbOSrKDGCfTjARP-FMOzuAeAJLo/viewform)) along with your meetup.com username. In other words, if your the name on your profile is "dataguy415," you will need to fill out the google doc. If you arrive to the talk and we haven't gotten this information, it is quite possible you will be denied entrance. We will NOT be following up with individual group members to track this info down - please make sure you take care of this.

Main Talk: Perils of Quantitative Investing

Speaker: David Andre, Ph.D (http://davidandre.com/) (Cerebellum Capital (http://www.cerebellumcapital.com/))

Abstract:

The field of quantitive investing is a fascinating one for machine with a wealth of data and opportunity. However, significant perils exist for the unaware, and many machine learning experts have been surprised at the difficulties in applying significant machine learning to finance. This talk will outline nine significant perils of machine learning in finance and present Cerebellum's primary methods for avoiding them.

Speaker bio:

Dr. David Andre is co-founder, CEO, and CTO of Cerebellum Capital, a hedge fund whose investment programs are based on a largely automated process for discovering financial trading strategies, evaluating strategies as to the likelihood they will generalize to unseen data, and combining these strategies to optimize the likelihood of significant steady return. David was instrumental in founding and building several companies, including BodyMedia, Inc, Blue Pumpkin Software, and Shinteki. In addition to holding numerous patents for his inventions, he is the author of more than 60 peer-reviewed publications in the areas of statistical machine learning, robotics, reinforcement learning, evolutionary computation, and parallel processing, as well as a book on automatic circuit design. Dr. Andre holds B.S. and B.A. degrees in Symbolic Systems and Psychology from Stanford University and a PhD in Electrical Engineering and Computer Science with a focus in Artificial Intelligence from U.C. Berkeley.

Lightning Talk: Distributed Machine Learning via Operator Splitting

Speaker: Brendan O'Donoghue (http://bodonoghue.org/) (Quantcast (https://www.quantcast.com/))

Tentative Schedule:

6:30-7:00 - socializing

7:00-7:20 - lightning talk

7:20-8:30 - main presentation

8:30-9:00 - socializing

Special thanks:

Quantcast (https://www.quantcast.com/) for hosting!