Viewing posts from March, 2017

Learning Reinforcement Learning

The inspiration for this site came from when I was watching a Machine Learning Course https://www.udacity.com/course/machine-learning--ud262 offered by Georgia Tech on Udacity. Here Michael Lipman and Charles Isbell mention a topic called mechanism design. The idea is to design the mechanism or incentives of the game to optimize the game's outcome.

Heart Disease

Heart disease is the leading cause of death worldwide. I found the dataset from the UCI repository which hosts a bunch of nice clean datasets, useful for learning and practicing data science. At first, I hoped to surpass the result of 78% accuracy which I did using commonly used classification algorithms.

Caught Illegal Immigrants and Naturalization

With all the current events with Trump as the president of the United States, it would be good to use data to fact check his statment and policies given his track record of alternative facts. Taking data from the US Border Patrol on the Customs and Border Protection webpage re-posted on Kaggle, a did a very rough analysis. Some data cleaning and manipulation using pandas allowed for a useful visualization of caught immigrants at the borders of the United States: