Joyce Robbins

interpretable machine learning packages/methods/graphing types/algorithms

How to do visualization summaries in R, and later something about D3 of java script.

Nothing at all. The lecturer taught graphics like math history. She gave many figures like mosaic plot, alluvial diagrams and etc.. However, when should I use those figures? Some figures seem very difficult to understand for normal people, which means they cannot understand what you draw.

This was probably my most helpful class this semester in terms of acquiring new skills that advanced my statistical ability. Learning how to take data and present it visually with tools such as R and D3 has been really valuable.

Lots of ways to make new graphs, the proper way to use GitHub, and D3. Good to learn the theory behind some of this stuff. Lots of hands-on exercises as well. Overall super useful!

It's a very useful subject matter, and the teacher made the material easy to follow

Introduction to various methods to make machine learning models more interpretable. PDP plots, SHAPly values, in addition to review of traditional ML models (Naive Bayes, linear regression, logistic regression, kNN, random forest, decision trees).

The final isn't the best way to test our knowledge - it seems kind of futile since you'd just google the things you don't know while coding (which is generally encouraged in the class) 1 Columbia University: Arts & Sciences Spring 2022 Course: STATGR5293_004_2022_1-TOPICSINMODERNSTATISTICS : STAT GR 5293 Statistical Graphics Spring 2022- STATGR5293_004_2022_1 - TOPICS IN MODERN STATISTICS Instructor: Joyce Robbins