William Lonon

Data Scientist specializing in machine learning, audio processing, and computational analytics. Building intelligent systems that extract meaning from complex data.

Python

ML/AI

Audio DSP

Optimization

Data Engineering

About

I'm William Lonon, a data scientist bridging the gap between audio engineering and machine learning. Completing my degree at the University of Arkansas with dual concentrations in computational analytics and music industry data, I build ML systems that extract meaning from sound—from drum classification models that parse raw audio into component samples, to predictive analytics for music production workflows.

My work spans geospatial crime prediction, optimization algorithms, and interactive data visualization. When I'm not training neural networks, I'm in the studio producing music, which gives me an intuitive understanding of the data I work with.

Audio ML: Drum Classification CRNN Deep learning model that processes raw audio and outputs sorted .wav files by drum type (kick, snare, hi-hat, etc.). Built with Python, TensorFlow, and custom audio preprocessing pipelines.

Geospatial Crime Prediction System ML system predicting crime likelihood across San Francisco using clustering and temporal pattern analysis. 8x more accurate than random baseline for hotspot detection.