• What we’ll do
Some light socializing, followed by a a few talks.
Michael Patterson will be giving his talk “What makes athletes popular: measuring bias on the NBA subreddit”
Michael Patterson is a former neuroscientist turned data scientist. In his spare time he does projects involving NLP and causal inference.
Social media is now where public opinion forms, and sports are an easily quantifiable way to analyze public opinion. In this project, I wanted to understand why certain athletes are popular on social media; specifically, how performance (e.g. points) and other demographic characteristics (age, race) influence popularity. To do this, I scraped millions of comments off reddit, and used sentiment analysis to measure how popular each athlete was. Then I used the econometric technique multivariate regression to understand which factors underly popularity. The basic results were that high scoring, young, and old athletes were popular. In terms of biases, I found that high scoring white players were popular, and there was a bias against black coaches.
Tim Crosley will be giving his talk “Validating Uncertain Data”
Tim Crosley is a software engineer at Domain Tools and maintainer of several notable Open Source Python projects.
Many of us, as software developers, value certainty and control. And those of us who are or work with Data Scientists, know well the value of consistent, well structured, and clean data. Indeed, many validation and data processing libraries and platforms assume complete control and specification of data is possible. However, while we may all want well-structured data, for a variety of reasons, we can not always demand it. In this talk, I will go over common strategies for dealing with unstructured data in Python, as well as their pitfalls. I will then introduce attendees to a new approach that side-steps common issues and enables responding to changes in source data quality in real-time.
• What to bring
Your wonderful self.
• Important to know
PuPPy has a Code of Conduct that applies to all PuPPy events and social spaces. To read the Code of Conduct go to http://www.pspython.com/pages/code-of-conduct/
Our after party today will be at Teku Tavern