Nicholas Colina

Nick graduated cum laude from the BS Physics program in 2017. His research was on finding simple ways to improve the outcomes of machine learning by deep neural networks. Also, as the HPC systems administrator of SanD, he has worked to integrate GPU capabilities into the laboratory's computing facility. He is now a data scientist for an AI consulting and development company in San Diego.

BS Thesis: 

Eigenspectra preservation and orthogonal weight matrix initialization in deep neural network models