Over the past year, I’ve had a successful collaboration with some great colleagues at the Applied Machine Learning Lab at Duke university.

The Cooper Pioneer published this article about closing the gender gap at Cooper Union, something that I think is extremely important for our future as an institution.

Are you bored on your break and want to find some new anime to watch? Why not try out Xiangling Kong and Venkat Kuruturi’s final project for statistical learning The data scraped for this rec system is uploaded to Kaggle in case you want to build your own:

Here is another outstanding project from my statistical learning course, this one formatted as a blog post. The project collected tweets about sexual violence, and used an unsupervised machine learning algorithm called latent Dirichlet allocation to discover hidden topic groupings. Its would be quite challenging for a human reader to parse huge amounts of text and organize them, so this is a really exciting area of research. There were some really surprising results that came out of this analysis, have a look. Please give Jess and Jason lots of encouragement so they keep working on this in the spring semester - this work has a lot of potential, hopefully this is just the start.

An example of some of the work my students did this semester. They scraped some data about car features and prices and built some models to predict the prices. They investigated what car features contribute most to the prices, and then, perhaps most interestingly, determined which cars are most overpriced based on predicted price vs actual price. Sound interesting? The data and code is open sourced now and up on Kaggle