Abstract: The rapid advances of sensing technologies (e.g., mobile devices, wearables, imaging, etc.) have produced immense amounts of data across different domains such as military, health, and engineering. The data is typically characterized by multiple modalities, high dimensionality, and large volume. Significant challenges exist to build novel statistical machine learning models that can transform the data to support decision making. In particular, there is need to balance data inclusivity and usability in a unified framework. Here, “inclusivity” means leveraging all the available data including those instances with incomplete information. “Usability” means automatically selecting the optimal subsets of data instances and features—reducing redundancy and noise to train predictive models efficiently and accurately. In this talk, Dr. Gaw will present research that develops a new semi-supervised learning model with simultaneous instance and feature selection (SSL-S2) to balance data inclusivity and usability in predictive modeling. Methodological development will be presented within the context of mobile health, i.e., the use of mobile devices and wearables to collect activity data of individuals for predicting their disease progression (e.g., individuals with Parkinson’s disease). The proposed SSL-S2 can also be used to address similar problems beyond health, such as in manufacturing fault detection and remote sensing domains. Additionally, Dr. Gaw will briefly introduce his other research works, including fusion of machine learning and mechanistic models, and interpretability/trustworthiness of deep learning models with application to imaging-based medical diagnosis.
Bio: Dr. Nathan Gaw is an assistant professor of Operations Research at Air Force Institute of Technology, Wright-Patterson AFB, Ohio, USA. His research develops new statistical machine learning algorithms to optimally fuse high-dimensional, heterogeneous, multi-modality data sources to support decision making in the military and healthcare settings (e.g., telemonitoring, diagnostics, combat recovery, etc.). He received his B.S.E. and M.S. in biomedical engineering and a Ph.D. in industrial engineering from Arizona State University (ASU), Tempe, AZ, USA, in 2013, 2014, and 2019, respectively. Dr. Gaw was a postdoctoral research fellow at the ASU-Mayo Clinic Center for Innovative Imaging, Tempe, AZ, USA, from 2019-2020, and a postdoctoral research fellow in the School of Industrial and Systems Engineering (ISyE) at Georgia Institute of Technology, Atlanta, GA, USA, from 2020-2021. He is also a member of INFORMS, IISE, and IEEE.