Visit
Fri 01/17
Ben Moseley headshot

Machine Learning for Faster Optimization

This talk explores the emerging area of algorithms with predictions, also known as learning-augmented algorithms. These methods incorporate machine-learned predictions into algorithmic design, allowing the algorithms to adapt to input distributions and achieve improved performance in runtime, space, or solution quality. The presentation will highlight recent advances in leveraging predictions to enhance the efficiency of optimization algorithms and dynamic graph algorithms.

A key focus will be on achieving “instance-optimal” performance—where algorithms excel when predictions are accurate—while ensuring graceful degradation when predictions are imperfect. Through examples such as bipartite matching, the talk will demonstrate the transformative potential of this approach to significantly improve algorithmic efficiency.

Speaker Bio

Ben Moseley is an Associate Professor of Operations Research at Carnegie Mellon University (CMU) and a consulting scientist at Relational AI. He earned his Ph.D. from the University of Illinois. His research has been recognized with numerous accolades, including Best Paper Awards at IPDPS (2015), SPAA (2013), and SODA (2010), as well as Oral Presentations at NeurIPS (2021, 2017) and Spotlight Papers at NeurIPS (2023, 2018).

Ben has been an Area Chair for ICML, ICLR, and NeurIPS annually since 2020 and has served on program committees for leading conferences, including IPCO (2025), SODA (2022, 2018), ESA (2025, 2017), and SPAA (2025, 2024, 2022, 2021, 2016). He was an Associate Editor for IEEE Transactions on Knowledge and Data Engineering (2018-2022) and has been an Associate Editor for Operations Research Letters since 2017.

His achievements include the NSF CAREER Award, two Google Research Faculty Awards, a Yahoo ACE Award, and an Infor Faculty Award. From 2018 to 2024, he held the Carnegie-Bosch Chair. Additionally, he was named a Top 50 Undergraduate Business Professor by Poets and Quants.

Ben’s research spans algorithms, machine learning, and discrete optimization, with a current focus on integrating machine learning robustly into decision-making processes.