Advances in machine learning, combinatorial optimization, and other types of mathematics, statistics, and computer science are increasingly being developed to address pressing problems in many disciplines, including systems biology, genomic biology, medicine, and social sciences. HPC enables these methods to scale to large datasets, but real world datasets are also highly heterogeneous. Hence, the computational challenges arising in this context go far beyond the “embarrassingly parallel” (i.e., a large number of relatively simple/cheap single analyses/runs to be done) and will require more HPC topics to be addressed in large-scale data analytics. We will discuss the question: What are the implications, needs, opportunities, and limitations?
This workshop focuses on applications where the dataset size requires new approaches and can benefit from HPC. These applications are also characterized by multiple types of mathematics and computer science, including combinatorial and graph-theoretic algorithms, so that this workshop complements the second workshop in this long program, which focuses on applications that are typically driven by ODEs and PDEs.
This workshop will include a poster session; a request for posters will be sent to registered participants in advance of the workshop.
Jennifer Chayes (Microsoft Research)
Vipin Kumar (University of Minnesota, Twin Cities)
Yann LeCun (New York University)
Tandy Warnow (University of Illinois at Urbana-Champaign)