A conversation with Thomas Wiecki on the use of probabilistic programming and machine learning in quant finance.
ConferencesModelingPredictive AnalyticsWebinarsposted by ODSC Team November 29, 2017 ODSC Team
Not surprisingly… hedge funds and especially quant funds are notorious for being secretive about the algorithms they employ to beat the market. A Boston based startup is taking a different approach.
Thomas Wiecki is the Head of Research at Quantopian, which hosts an open source platform that allows anyone on the web to build, share, backtest, and trade quant finance models. Quantopian employs some of the best performing strategies to run a crowdsourced hedge fund.
In this conversation Thomas explains the advantages of employing probabilistic programming over more traditional methods. He touches on several benefits including flexible time series models and dealing with uncertainty. Insightful, he delves into a number of additional topics including why machine learning may not always be ideal for quant finance modeling, backtesting in relation to in-sample and out-of-sample data, distributions to employ, modeling volatility, and more.