Coursework

Anyone can learn how to code a regression model; the real challenge is understanding the math that lies under the hood. This class is notorious for being one of the most difficult in the major. Grasping the complex proofs and theorems behind linear models took grit and determination, but it made me a much stronger statistician and data scientist.

Simple linear regression can be powerful, but only when used correctly. This class taught me the importance of experiment design, understanding how data is collected and how it should be analyzed. Messing up a block design can lead to disastrously invalid conclusions; I learned the hard way.

How can you tell species of penguin apart? (Besides looking at them, of course.) The answer is regression analysis! There is something intensely satisfying about creating and then using a decision tree to classify a penguin as Gentoo or a Star Wars character as a droid.

This class featured one of my favorite homework assignments across all my classes at UCLA. We were tasked with creating an MLP (multi-layer perceptron) to distinguish handwritten drawings of different digits, specifically 3 vs 5 and 4 vs 9. After lots of testing and tweaking, my MLP achieved an AUC of 0.9996 for the digits 3 and 5. R code had never felt so powerful!

I've been using simple Monte Carlo Methods in sports simulation for over a decade, so learning and grasping the elaborate theory and extensive applications of MCMC was exhilarating!

Spatial data is everywhere! This was one of my favorite classes at UCLA because it gave me a greater appreciation for geostatistics and the methods we use to analyze spatial data. I particularly enjoyed working with a California housing prices dataset, using kriging to interpolate property values across the state.

This class opened my eyes to the expansive world of finance and investment. I have always been interested in stock data, so getting my hands dirty with real-world market data was both illuminating and invigorating.