All events are in the Data Science Initiative seminar room, 360 Shields Library from 12 - 1.30pm on Friday, unless otherwise noted.
Please suggest topics you would like to have a workshop on. Send an email to email@example.com.
It is relatively easy to write R code quickly, but harder to write quick R code. We'll discuss some general strategies for writing faster R code and then explore R tools for finding bottlenecks in code and how to make these faster. We'll use one or two in-depth case studies to see the process and reasoning you would use on code. We'll also look at alternative approaches to making code faster by integrating other languages, e.g., the shell, SQL and databases, and C and C++. The intent is to show that these are quite accessible.
Overfitting and Surrogate Losses
Non-Parametric Methods and the Kernel Trick
In this workshop, we'll discuss best practices for displaying different types of static data. This is about designing data visualizations rather than the software and instructions we use to actually render them. We'll cover some basic principles to keep in mind, explore different types of plots, critique some good and bad examples of displaying data and how to think about creating effective plots.