Denoised hyperspectral datasets of hybrid perovskite materials in Python, and used Scikit-learn and Nimfa to build unsupervised Non-negative Matrix Factorization models to extract image and spectral features
ML analysis of HDF5 data (VAE)
Trained, validated, and analyzed unsupervised and semi-supervised variational autoencoder models of hyperspectral datasets of energy materials and of artificial datasets in Pytorch, demonstrating relationships between image and spectral features
Full stack personal blog
Full stack web app built for Ruby M. using React, Node.js, Express.js, and MySQL, deployed on a Digital Ocean droplet using an Nginx reverse-proxy server
Sorting algorithm benchmark
Compared 8 sorting algorithms implemented and tested in 4 languages: Python, Java, C++, and JavaScript; deployed the results to a static web page, visualized using Chart.js
Searching algorithm benchmark
Compared 3 searching algorithms implemented and tested in 4 languages: Python, Java, C++, and JavaScript; created data visualizations using Chart.js
Telegram Chat CC Parser
Open source intelligence tool built in Python to parse JSON files exported from Telegram channels for credit/debit card information, built to aid response of leaked sensitive information and combat financial fraud
