Gautam Dasarathy

Gautam Dasarathy is an Associate Professor in the School of Electrical, Computer, and Energy Engineering at Arizona State University, where he also holds a courtesy appointment in the School of Computing and Augmented Intelligence. Additionally, he serves as an Amazon Scholar, working on machine learning and optimization problems relevant to Amazon Last Mile. His academic journey spans prestigious institutions including Rice University, Carnegie Mellon University, and the University of Wisconsin-Madison.

Dasarathy's educational background reflects a strong foundation in electrical engineering and machine learning:

• Ph.D. in Electrical Engineering, University of Wisconsin-Madison (2014)
• M.S. in Electrical Engineering, University of Wisconsin-Madison (2010)
• B.Tech. in Electronics and Communication Engineering, VIT University, India (2008)

Dasarathy's research lies at the intersection of machine learning, statistics, information processing, and networked systems. He specializes in developing data- and compute-efficient learning algorithms for resource-constrained environments, with a particular focus on interactive learning where algorithms decide what data to collect next. His work frequently leverages structural constraints such as graphs, manifolds, or physical laws to inform both inference and data acquisition.

His expertise spans multiple domains including Machine Learning, Network Science, Phylogenetics, Signal Processing and Communications, Statistics, and Systems and Control Theory. Recent applications of his research include power grid monitoring, neuroscience, meta-science, circuit design, and epidemiological forecasting.

Dasarathy's recent publications demonstrate a consistent focus on graph-based learning, active learning methodologies, and resource-constrained machine learning. His work spans theoretical foundations in statistical learning and practical applications across diverse domains. A notable trend is the integration of domain-specific constraints (particularly graph structures) into learning algorithms to improve efficiency and accuracy. His research increasingly addresses challenges in federated learning, Bayesian optimization, and meta-science applications.

Dasarathy has received numerous prestigious awards recognizing both his research and teaching excellence:

• 2024 Top 5% Teaching Award from ASU's Fulton Schools of Engineering
• 2022 IEEE Transaction on Haptics Best Application Paper Award
• Distinguished Alumni Award (Academics) from VIT University
• NSF CAREER Award for research on graph structure learning
• AISTATS 2021 Oral Paper (top ~3% of submissions)
• Multiple papers accepted to top-tier conferences including NeurIPS, ICASSP, and ECCV

Dasarathy actively mentors graduate students, with Parth Thaker recently completing his thesis on bandits, interactive learning, multi-agent systems, and nonconvex optimization. His research program is supported by significant funding from multiple federal agencies. He serves as PI or co-PI on grants from NSF (including CAREER, RAPID, and PIPP programs), DARPA (Geometries of Learning program), ONR (Active Meta Learning), and NIH (Graphical Model Selection from Partial Measurements). His collaborative projects span disciplines from power grid monitoring to epidemiological forecasting, demonstrating the broad applicability of his methodological contributions.

Dasarathy leads a research group focused on machine learning and networked systems at ASU. His team works on both theoretical foundations and practical applications of learning algorithms. He is part of several interdisciplinary initiatives at ASU, including collaborations with the Learning and Teaching Hub on AI in education. As an Amazon Scholar, he bridges academic research with industry applications, particularly in last-mile delivery optimization.