Data Science Portfolio

Tweet Prediction

The goal of this project was to classify twitter feeds in AM or PM, Do Sentiment analysis of ten news organizations, and Use Monte Carlo Simulations (pymc3) to estimate the difference in tone between different classes.

View Project

US Department of Education

In this project, I assembled a dataset for investigating colleges and universities through an API from the US Government, Built a model with pymc3 to estimate the difference in tuition between Private nonprofit and Private for profit schools, Built a Neural Network Classification Model using Keras.

View Project

Ames Housing & Kaggle Competition

The Ames Housing Dataset is an exceptionally detailed and robust dataset with over 70 columns of different features relating to houses. I built models to predict the following: The price of a house at sale (regression) Whether a house sale was abnormal or not (classification)!

View Project

Drug Addiction

I analyzed a 2012 survey on drug use from the federal Substance Abuse and Mental Health Data Archive (SAMHDA) to see what drugs baby boomers are taking now. Are their patterns of use different from other age groups? How similar are people within the baby-boomer cohort1 when it comes to drug use? I also created a testable Hypothesis about two of the most use drugs.

View Project