Projects

In this project (my capstone at UCR MSDS Program), I go through a toy example of trasnfer training the huggingface BERT large language model to perform sentiment analysis on social media comments and compare its performance against a Naive Bayes Classifier, which is a much smaller model and much easier to implement.

Naive Bayes Classifier From Scratch

The aim of this paper and project is to recreate my own implementation of the Naive Bayes Classifier. The paper then compares the performance of my implementation against the Scikit-learn implementation.
Code here

Classifying Shopper Intention

The aim of this project is to try to predict whether an online shopper will go through with a purchase or exit the webstie they are on. The data is taken from a dataset in the UCI Machine Learning Library.

Predicting Carbon Export Using Satellite Data to Estimate Mean Planktonic Size and Sinking Speed

This project was my senior thesis that I completed a little after 1 year of research in my lab. Our lab is a physical oceanography lab, so my thesis focused on using remote sensing data to estimate a measurement of plankton, combined with in-situ data to create a carbon export model using a modified sinking speed model of diatoms. More can be found in the thesis!

Comparisons of Physical Properties of Ocean Subsections of Similar Latitudes to In Situ Data

This was a analysis of thermophysical properties of ocean subsections of similar latitude. This was a paper that served as a prerequisite for my senior thesis.

Profiling Potential Mortgage Customers Using Pennsylvania Census Data

This is a project I worked on for a Data Analysis class as part of my Data Science Certificate at UC Irvine. This project centers around using KNIME to create a workflow for importing data in KNIME, cleaning the data, splitting the data into training / test sets, and creating a model to predict final outcomes of whether a customer purchases a mortgage.

PumpScore

Automatic Score Detection and Recognition by Computer Vision using OpenCV and Tesseract. This project focuses on automatically extracting data from a score screen in the popular dance / rhythm game "Pump It Up".

Currently working on formatting and uploading more of my projects in a site-friendly format!