Speaker: Olgica Milenkovic, University of Illinois at Urbana-Champaign
Abstract: The emergence of massive amounts of complex and heterogeneous data from social networks, molecular biology and physics has set the stage for a new scientific revolution. As it is no longer possible to adapt existing computational, storage and learning paradigms to effectively utilize such data, new interdisciplinary research approaches are needed to address existing processing challenges and bottlenecks.
In this short overview talk, we describe several new problems at the interface of machine learning, graph theory and error-control coding arising in the context of macromolecular storage system design and implementation, large-scale network data analysis, and distributed data storage and deduplication. We also illustrate how solutions to these problems may be used for practical data analysis.