Michael Fang

Reveal Contact Info

PhD Student

Physics

DeWeese Lab

Current Research

Currently I’m working on two projects, optical neural networks (ONN) and autonomous experimentation through surrogate modeling.

Collaborating with Intel, I’m modeling and simulating ONNs — neural network physically implemented with photonic components on silicon photonics. Specifically, I’m interested in neural networks with noisy weights during inference. This is important for ONNs as all analog computation devices are suspect to the effects of noise.

Working with the Berkeley Synchrotron Infrared Structured Biology (BSISB) at Berkeley Lab, I’m also implementing autonomous experimentation for Fourier-transform IR (FTIR) spectromicroscopy. In FTIR imaging, spectral response of individual pixels are collected. To improve beyond scanning in a uniform grid, I am using surrogate modeling to estimate the most optimal location for subsequent measurements.

Background

I grew up and went to college in Toronto where I often go back to visit friends and family when I miss snow a bit too much. After finishing undergrad, I came to Berkeley to start my PhD in physics. Before arriving at the Redwood Center, I did some work in condensed matter and quantum optics. However, eventually, I found interest in exploring problems outside of the traditional fields of physics. When I’m not working, I enjoy traveling and photography to maintain my sanity.