Projects

My newest project I am working on is training a neural network to not only give song suggestions, but build playlists and even DJ sets by user inputted information. I found this project to be really exciting as it combines coding along with my music interests! I hope to have this fully developed into a website and application that people and DJ's turn to for suggestions on songs or sets.

This project involved analyzing a dataset of potential customers for a B2B product. EDA was done to identify a game-plan and then different ML models were trained to help the sales team identify the hottest leads that were likely to convert. This model had 97% precision for positive targets which meant 97% of all future customers were identified.

In order to create this credit risk model, 2 different datasets were used since each had unique variables and a KNN imputer was used to fill in the missing data. After training the models and identifying the final one, the feature importances (correlation to target) were set to an equation to calculate a credit risk score for potential applicants. This risk score was then placed into Streamlit and then into AWS EC2 for people (my friends & professor) to enter their information and see their loan amount and interest rate.

When I began creating music I discovered sample packs, which contain hundreds if not thousands of different audio files. I chose to create an automated file that would organize all of these for me from my downloads by redirecting their paths into my samples folder. Different folders include kicks, snares, risers, sound fx and plenty more if you want to run this to help organize your music samples folder!

My University projects included technologies like Python, R, SQL, Tableau and Qlik. Different types of analysis, machine learning methods, and visualizations were created throughout this program. In order to respect their academic policy, this repository is private but I can grant access to it if needed.