Josh Solis

We built a deep learning model using several LSTM layers to train on classical snippets and generate new melodies based on composers’ melodies. The web app includes use of AWS Elastic Beanstalk and Netlify for hosting. Melody Interpolator [Read More]

Principal component analysis was used as the primary data science tool in order to find songs similar to the user’s tastes. We used a Spotify dataset with information on songs like “danceability, energy, mode, key, pitch, tempo, timbre, and so much more, Queue searches through thousands of songs and their... [Read More]

We used time series data analysis on time series data for deforestation in countries around the world. Python tools like matplotlib, pandas, and statmodels were crucial in the proper analysis of the data given. Link to project: Deforestation Dashboard. [Read More]

My project on gun violence data was insightful, and summarized on an interactive plotly map of the USA. I found that Chicago was the most dangerous city in the country when it comes to guns. Chicago alone has more gun victims than all other states. I wrote about my findings... [Read More]

I engineered features, tried different methods, and finetuned models. My best one is written about here. [Read More]