Kickstarter Success Predictor is a web app that allows a user to quickly determine if their Kickstarter Campaign will be successful based on data from prior campaigns. This web app is built on FastAPI and deployed on Heroku. The app uses an XGBoost model with certain optimized hyperparameters to generate a prediction.
Spotify Song Suggester is a web app that helps users find and visualize songs that fit their personal taste. This app uses NLP and the KNearest Neighbors algorithm to take an input song and find your top 5 similar songs.
Twitoff is a web app that utilizes the twitter api to take an input "tweet" and return who is more likely to tweet it. This web app is built on Flask and deployed on Heroku. The app uses NLP to vectorize tweets then uses logistic Regression to make a prediction.
Housing price prediction is a blog I wrote about the use of ensemble methods to predict housing prices and building type. This blog was published in Geek Culture and showcases various models to find the best model for each problem.
Is Data our Future is a blog I wrote about the technological advances we have made resulting in an abundance of data. In this blog I utilize the t-test, correlation coefficient, pivot table, heat map, and more to analyze questions about Data Analyst and Data Scientist job listings.