Gautam Dasarathy

Rice University, Arizona State University

Papers

Total Citations

135

H-Index

About

Gautam Dasarathy is a leading researcher in the field of Bayesian optimization, with a particular focus on developing algorithms for expensive, black-box functions. His major contributions center on multi-fidelity optimization and high-dimensional Bayesian optimization, addressing the critical challenge of efficiently optimizing functions when evaluations are costly but cheaper approximations are available. His seminal 2016 paper, "Gaussian Process Bandit Optimisation with Multi-fidelity Evaluations" (68 citations), introduced a principled framework for leveraging multiple information sources of varying fidelity, dramatically reducing the number of expensive evaluations needed. This work has been foundational for applications in scientific computing, engineering design, and machine learning hyperparameter tuning. His 2021 survey, "Bayesian Optimization in High-Dimensional Spaces" (60 citations), has become an essential reference for researchers tackling the curse of dimensionality in BO. Dasarathy's work is distinguished by its rigorous theoretical foundations and practical impact, enabling optimization in domains where traditional methods are infeasible. His research continues to shape how scientists and engineers approach complex, resource-intensive optimization problems.