I implemented a gaussian matrix factorization model and used MAP estimation on the MovieLens dataset to predict the rating a user would give to an unseen movie. All algorithms were written from scratch.
Read more about the data, model, data preparation, and practical learning during MAP estimation in 6701_HW2.pdf.
The code is in the jupyter notebook.
Code was shared with over 100 students in a PhD Level Statistics and Machine Learning Course as a representative solution.
Course: Foundations of Graphical Models by David Blei, Columbia University, NY.