Exploratory Autoencoder

For this project, I utilized generative AI ( variational autoencoder) and "visualization" tools (PCA and t-SNE) to generate two-dimensional encodings of 1000+ theoretically simulated x-ray spectra. These encodings produced higher accuracy when classifying these spectra with supervised machine learning than using the spectra themselves. This paper is my highest cited one, and it was recognized by PCCP as one of their HOT articles of the year.