Benjamin Goodrich

How to use R; Bayesian statistics; rstanarm

The class is roughly divided into two parts: 1. programming best practices, exploratory data analysis (EDA), and unsupervised learning 2. supervised learning including regression and classification methods In the first part of the course we will focus writing R programs in the context of simulations, data wrangling, and EDA. Unsupervised learning is focused on problems where the outcome variable is not known and the goal of the analysis is to find hidden structure in data such as different market segments from buying patterns or human population structure from genetic data. Supervised learning deals with prediction problems where the outcome variable is known such as predicting the price of a house in a certain neighborhood or an outcome of a congressional race.

Introduction to R (dplyr, ggplot2), natural language processing (text and sentiment analysis), unsupervised learning, and supervised learning (linear regression, classification, non-linear models, deep learning and neural networks)

Learned supervised learning concepts, unsupervised How to wrangle and analyze data in R including modern best practices Everything is done as closely as possible in a social science context Social considerations in ML/AI (ethics)

I learned a bit about Bayesianism, a bit about the short comings of frequentist approaches, and a bit about RStanarm.

The course starts with an introduction to R, which seems well paced initially. However, post midterm, the class begins to get steeply uphill, with a new topic introduced almost every week. That doesn't allow students an opportunity to really understand a topic. Generally, it feels like the course is too broad, but with very little depth in any topic. The workload and assignments are nice, but the exams are harsh.

This course doesn't make much sense to me. Although the professor said that the approach to this course is very radical compared to other schools, I don't think that "radical" means "good" in this case. On the contrary, this course is not useful for students without a basic knowledge of frequency statistics, as it does not properly introduce the basic ideas of statistics and probability. It's also not very helpful for those with some experience since it's not about Bayesian statistics but just about some random comparisons between Bayesian and "Frequentist" (whatever it means) stats. The essence of the course is not clear to me, I do not feel that I am learning anything from it.

Data Mining: R (dplyr + ggplot2), _some_ linear algebra, how to google your way around questions and issues, some unsupervised and supervised learning.

I would rather suggest taking some introductory course on econometrics or linear regression. I do not think that anyone should start learning stats with Bayesian perspective on it.

lots of R. how to code in R. how to do things in R. Some linear algebra for regression. Some analysis techniques. supervised learning. a lot of data wrangling skills.