I am an Assistant Professor in the AMS department at Colorado School of Mines, where I am a member of the Mines Optimization and Deep Learning group (MODL). I received my PhD from the University of Georgia in May 2019, under the supervision of Ming-Jun Lai. Prior to that I was an undergraduate at the University of Cape Town, from where I also received my Master’s degree. In between Cape Town and Athens, Georgia I spent a semester at the University of Bayreuth. My work lies in the intersection of machine learning, signal processing and optimization. More specifically, my research interests are
- Derivative-Free optimization (DFO), i.e. finding a minimizer of a function whose gradients are inaccessible. I am interested in developing methods for high-dimensional DFO and applying them to problems in machine learning, for example studying adversarial attacks on image classifiers or fine-tuning large language models. For a representative paper, see here.
- Learning to Optimize. I focus on combining classical optimization tools with modern deep learning to solve optimization problems where some parameters are unobserved. See here and here.
- Topological and geometric methods in data science. In particular, geometry-inspired metrics on data sets and geometric approaches to adversarial training. See here and here.
You can find an up-to-date list of publications on my Google Scholar page.
I enjoy teaching upper and lower division courses, and strive to make my classrooms a welcoming space for all. You can find more information on my teaching page. This semester I am teaching Math 551: Computational Linear Algebra.
You can reach me at d[my surname]@mines.edu. My physical address is 1500 Illinois Street, Golden, CO, 80401.