Wesley J. Maddox

Papers

Total Citations

H-Index

About

Wesley J. Maddox is a rising researcher whose work lies at the intersection of Bayesian optimization, Gaussian processes, and decision-making under uncertainty. His research focuses on making probabilistic machine learning methods practical for high-stakes, real-world applications where data is scarce and decisions must be made sequentially. Maddox has made key contributions to scaling Bayesian optimization to handle high-dimensional outputs, enabling more efficient optimization of complex, multi-objective problems in fields like scientific discovery and engineering design. His work on conditioning sparse variational Gaussian processes for online decision-making addresses a critical bottleneck: the computational cost of traditional Gaussian processes, which scale quadratically with data. By developing methods that maintain principled uncertainty estimates while reducing computational demands, Maddox is helping bridge the gap between elegant Bayesian theory and deployable algorithms. Though early in his career, his papers—including his 2021 work on high-dimensional Bayesian optimization (9 citations) and his 2021 paper on sparse GPs for online learning (5 citations)—demonstrate a clear trajectory of impact. Maddox’s research is particularly relevant for students and practitioners interested in sample-efficient learning, active learning, and Bayesian methods for autonomous decision-making systems.