Benjamin Marlin

Professor
 
Phone: (413) 545-4493
A355 LGRC

I am a professor in the Manning College of Information and Computer Sciences at the University of Massachusetts Amherst where I direct the Robust and Efficient Machine Learning  (REML) Lab. I was previously a fellow of both the Pacific Institute for the Mathematical Sciences and the Killam Trusts at the University of British Columbia where I was based in the Laboratory for Computational Intelligence in the Department of Computer Science. I completed my PhD in machine learning in the Department of Computer Science at the University of Toronto.

Prospective Students

  • PhD Applicants: I expect to have two opening for new PhD students in my lab in 2025. The first position is in machine learning methods for spatiotemporal reasoning starting in January 2025 or later, and the second is in machine learning for mobile health starting in Fall 2025. See here for details and how to get in touch regarding these positions.
  • MS Students: I regularly supervise research-based MS independent studies in the spring semester each year for top-performing students in COMPSCI 689. Please reach out after final grades post to discuss opportunities.
  • Undergraduate Students: I regularly supervise honors theses for students interested in machine learning. Students need to be taking COMPSCI 589 in the semester they are looking to submit their 499Y proposal. Please reach out around mid-semester to discuss opportunities.

Research Interests

From a methodological perspective, my current research focuses on the development of machine learning models and algorithms that are both robust and efficient. The REML Lab studies multiple aspects of robustness including robustness to uncertainty and missing data, as well as multiple aspects of efficiency including data efficiency, computational scalability, and communication efficiency. The Lab’s foundations are in probabilistic machine learning and deep learning. Recent research topics include modeling sparse and irregularly sampled time series, methods for learning from scarce time series data data (including domain adaptation, active learning and Bayesian deep learning), and optimizing models for resource-constrained and real-time deployment. See the REML Lab web page for details on lab members, current and completed projects and publications.

Research Collaborations

My research is informed by multiple real-world application domains and machine learning deployment contexts including clinical and mobile health, embedded systems, and the Internet of Things. My lab has been supported by multiple large-scale collaborative research grants including the MD2k Center, the mDOT Center, MassAITC, and the IoBT Collaborative Research Alliance. Through these and other projects, my research group collaborates widely with behavioral scientists, health researchers, statisticians, engineers and computer scientists from UMass Amherst and across the country.

Affiliations