Analyzed if there is any bias within a specific hate speech algorithm by designing a multi-classification model using various NLP techniques with an accuracy of 84%, precision, and recall scores above 90%.

With a team of data scientists, we did a deep analysis of hate crime prosecution rates by building a predictive classification model to determine if a hate crime case will reach prosecution.

Cleaned, examined, and recommended solutions from SAT/ACT datasets to help an admission board increase their African American student population using exploratory data analysis, python, and other analytical tools.

Acquired data using Pushshift's API to achieve various NLP techniques, including Term Frequency-Inverse Document Frequency or CountVectorizer, and designed multiple models such as Multinomial Naive Bayes and Logistic Regression to classify similar subreddits from one another.

Created linear regression models to predict sale prices of homes in Ames, Iowa. Used various feature selection, engineering, and regularization methods that accounted for 85% of the variability in housing prices.