What I've Worked On

My portfolio website built using Next.js, React, and Bootstrap. Animations were created using Framer Motion

Collaborated with a team of four students to design and implement a convolutional neural network model. Successfully utilized image processing and machine learning techniques to classify Pokémon sprites within a set of images, achieving up to 86% accuracy in classifying sprites from battle images.

In this project, we aim to predict the points spread of an NFL game using many different factors, including home and away team records and average per game metrics. To make these predictions effectively, we tested various regression models, including linear regression, ridge regression, lasso regression, random forest regression, voting regression, support vector regression, and a neural network. We believe training all of these various models will give us better insight into where we can improve model performance by introducing a different feature set, as well as being able to compare the accuracy of a purely linear model with a non-linear model.

In this project, we propose a method to detect toxic statements and identify specific toxic words within these statements. Additionally, we developed a method to completely censor these harmful words to prevent their negative impact in online spaces.