Multi-Data (MData) Lab: Modeling the real world heterogeneity
Data in the real world is seldom compartmentalized as is the model in traditional approaches. For example, data from one phenomenon at a location (e.g. health disparities) often intersects with or relates to or impacts other phenomenon (e.g. presence of food deserts, percentage of teen pregnancies) in the same location. So this begs the question: why do we study these phenomena in separate silos in algorithmic pattern discovery? Tackling this question and crossing over the data silos into a messy heterogeneous world is at the heart of the research in the Multi-Data Lab.
The MData lab has an active grant portfolio of over $15Million across multiple NSF projects leading to major societal impacts. Multi-domain relevance is evident in all types of social impact application areas such as climate change, and misinformation detection. In addition, data often is multi-resolution, collected with multi-modal mechanisms, multi-dimensional and originates from multiple disciplines. This is the true essence of this lab in dealing with the heterogeneity across multiple aspects of data, working across disciplines leading to social impact.
MData- copyright 2021