Oct'18-Using automatic derivatives to predict a ping pong match winner(Alex Fun)
Details
We look forward to seeing you at our October meetup
Arrive from 5:45pm for a 6pm talk.
Using automatic derivatives to predict a ping pong match winner.
TALK OUTLINE:
In a recent and popular Cross Validated post, the following question was asked:
“I bet with my colleague that I will beat him in fifty consecutive ping pong games. So far I have won 15, what are my chances of winning the next 35 games?”
-- from https://stats.stackexchange.com/questions/329521/
In this talk, Alex presents his own playful take on this problem. He proposes the following data generation process for the score-line of each game: The original poster is a far superior player who still wishes to make the game fun for their opponent (they are colleagues after all). This leads to a regression problem that cannot be fit using standard regression packages. Instead, he will demonstrate how to, and why you should (hint: it's magic!), use packages like TMB, rstan, and tensorflow, that provide automatic differentiation in R, to find maximum likelihood estimates for the regression coefficients. In addition, he will touch upon Rcpp, neural networks, and show you lots of beautiful maths and graphs.
BIO:
Alex is very happy to be able to return to give another talk at SURF. By day, he works with sports data, by night, he likes bicycles, coffee, wine, cooking, and learning more about statistics and machine learning.