Rate My Professor Peter Loxley

Dr. Loxley's postdoctoral career commenced with a position in mathematical neuroscience at the School of Physics, University of Sydney, where he explored nonlinear neural dynamics and proposed a model in which neural competition triggers a nonlinear wave instability to explain the visual percept of binocular rivalry. He then served as a postdoctoral researcher at the Center for Nonlinear Studies at Los Alamos National Laboratory, developing a statistical model of sensory processing akin to independent component analysis that accounts for redundancy reduction by simple cells in the primary visual cortex, and studying bi-stable steady states in two-dimensional turbulence through maximum entropy states in the point-vortex model, motivated by ocean current dynamics like the Kuroshio. Before joining the University of New England, he was an adjunct lecturer at the University of New Mexico in Los Alamos, instructing first-year mathematics and computer science courses. His research interests include statistical physics, optimal control, applications of information theory and coding, probabilistic models in machine learning, algorithms and computation, nonlinear dynamics, and pattern formation. He teaches units such as Discrete Mathematics, Data Structures and Algorithms, Algorithms in Machine Learning, and Reinforcement Learning. Key publications feature 'A Dynamic Programming Algorithm for Finding an Optimal Strategy Set in Sequential Multisensory Integration' (Entropy, 2023, co-authored with Ka-Wai Cheung) and 'A sparse code increases the speed and efficiency of neuro-prosthetic decoding' (Neurocomputing, 2021).