Using data visualization to find patterns in multidimensional data
Abstract: While machine learning methods have made great strides in predictive analytics, there are many components of data science that still require human intervention. In particular, people are great at finding visual patterns in data. John Tukey was talking about exploratory data analysis in the 1970s, but advances in computer graphics have given us additional powers. I'll demonstrate methods for finding patterns in high-dimensional data, including the generalized pairs plot, the Grand Tour, and the lineup protocol for graphical inference. R code will be provided in the slides, and attendees will be welcome to follow along and try things out on their own.