Project Guten-Bag-of-Words

project blog post

Post thumbnail
Post thumbnail
Project Gutenberg is a wonderful service providing free access to books with expired copyrights. However, very little information beyond the author and title are typically available through the site. This motivated me to use unsupervised machine learning in the form of natural language processing (NLP), dimensionality reduction, and clustering to... [Read More]

Some ideas

brainstorming blog post

Post thumbnail
Post thumbnail
Some ideas about my current project: I’m currently using Project Gutenberg for unsupervised learning. I’ve performed topic modeling with LDA and NMF using a chaotic assortment of over 1700 books, mostly fiction and science fiction. [Read More]

Predicting patient early readmission to the hospital

project blog post

Post thumbnail
Post thumbnail
I used several supervised machine learning algorithms to predict early hospital readmissions based on a dataset available from UCI here originally part of this study. Additionally, I performed a cost benefit analysis to optimize a threshold for recall and precision. Finally, I made a simple Flask prediction app as well.... [Read More]

Steam Webscraping and Linear Regression Project

project blog post

Post thumbnail
Post thumbnail
Steam is a popular online store for computer video games. I scraped Steam and steamDB data for every video game available for purchase through Steam and then performed linear regression to predict the number of concurrent users. This was done using Selenium and Python. Many Python packages were used including... [Read More]