Yerem Istanboulian

Classifying Amazon Reviews using Naive Bayes

Building a Naive Bayes Model to classify video game reviews as positive or negative

This project started as a coding assignment for my senior year Artificial Intelligence course at UC Santa Barbara. We were assigned to write code to run the Naive Bayes algorithm from scratch in Python without using any external packages like sci-kit-learn. We were given a training set and a testing... [Read More]

Predicting Pitchers’ ERA Using Skill-Based Measures

Building a Linear Model to predict ERA

Baseball is a sport that has always celebrated the use of statistics to compare players and teams. Coaches and managers have used data to determine how to improve their team for several decades and even fans of baseball often have an acute knowledge of baseball statistics. The most commonly used... [Read More]

The Future of the NBA Three-Point Shot

UCSB Data Science Club project that forecasts the future trends for the NBA three point shot

This project was conceived during a Data Science Club meeting at UC Santa Barbara by five students who shared two common interests: Data and the NBA. Given the recent rise in three point attempts in the NBA, we were all interested in forecasting how long this trend would continue. To... [Read More]