About the Lab
The REML Lab focuses on making machine learning models and algorithms more robust and efficient for use in challenging deployment scenarios. The Lab conducts research on multiple aspects of robustness including robustness to noise, missing data, low labeled data volumes, and the potential for out-of-distribution and adversarial examples. The Lab has a particular emphasis on solving these problems in the context of accuracy-speed-storage-communication trade-offs motivated by embedded, cloud-assisted and and other real-time deployment scenarios.
The lab pursues research on a variety of solutions to these problems founded on principles of probabilistic machine learning. Current research includes fusions of probabilistic and deep learning models for learning from incomplete and irregularly sampled data, the study of active learning and zero-shot learning for improving predictive performance in the low labeled data setting, and the study of computationally efficient Bayesian deep learning methods for deployment in resource constrained settings.
The Lab’s research is informed by multiple real-world applications domains and machine learning deployment contexts including clinical and mobile health, embedded systems, and the Internet of Things. The lab is deeply engaged in large-scale interdisciplinary research projects and collaborates with computer scientists, engineers, statisticians, behavioral scientists and clinicians locally and across the country.
The lab’s research has been supported by a number of grants, many of which are large-scale national collaborations including the NIH-funded MD2K Center and follow-on projects, the NIH-funded ILHBN Network, and the ARL-funded IoBT Collaborative Research Alliance.
|[2018-2022] Operationalizing Behavioral Theory for mHealth: Dynamics, Context, and Personalization (with Donna Spruijt-Metz, USC and Predrag Klasnja, U. Michigan). See the ILHBN Network website.|
|[2018-2021] mResearch: A Platform for Reproducible and Extensible Mobile Sensor Big Data Research (with Santosh Kumar, U. Memphis, PI).|
|[2017-2020] Enhancing Context-Awareness and Personalization for Intensively Adaptive Smoking Cessation Messaging Interventions. See NSF award listing.|
|[2017-2022] Alliance for IoBT Research on Evolving Intelligent Goal-driven Networks (IoBT-REIGN) (with Prashant Shenoy, UMass PI. UIUC prime to ARL.). See ARL and UMass Amherst press releases, and the IoBT website.|
|[2014-2018] Center of Excellence for Mobile Sensor Data to Knowledge (with Santosh Kumar, U. Memphis, PI). See center website.|
|[2017-2018] mPerf: A Theory-driven Approach to Model and Predict Everyday Job Performance Using Mobile Sensors (with Deepak Ganesan, UMass PI. U. Memphis prime to IARPA).|
|[2014-2019]. NSF CAREER: Machine Learning for Complex Health Data Analytics.|
|[2013-2016] Accurate and Computationally Efficient Predictors of Java Memory Resource Consumption (with Eliot Moss, PI).|
|[2012-2015] SensEye: An Architecture for Ubiquitous, Real-Time Visual Context Sensing and Inference (with Deepak Ganesan, PI).|
|[2012-2015] Patient Experience Recommender System for Persuasive Communication Tailoring (with Tom Houston, UMMS, PI).|
|[2012-2014] Foresight and Understanding from Scientific Exposition (With Andrew McCallum, PI and Raytheon BBN Technologies)|