报告题目：High-Performance Evolutionary Computation for Scalable Spatial Optimization
报告日期及时间：2017年11月15号上午10：00
报告地点： B404
报告人：Yan Liu
报告人国籍： 美国
报告人单位：UIUC
报告人简介：
Yan Liu is Director of Technical Program at the CyberGIS Center for Advance Digital and Spatial Studies and Senior Research Programmer at the National Center for Supercomputing Applicationsand the Department of Geography and Geographic Information Science atthe University of Illinois at Urbana-Champaign. He obtained his PhD in Informatics from the University of Illinois at Urbana-Champaign, MS from the University of Iowa, and BS and ME fromWuhan University. His primary interest is in the applied research in high-performanceand scalable algorithms and software, with a particular interest in domain-specific high-performance heuristic algorithms. At the CyberGIS Center, he is responsible for thesoftware architecting and development of CyberGIS core capabilities and funded projects.He is also a staff scientist in XSEDE, a leading national cyberinfrastructure in U.S., to provide advanced scientific computing consulting. He has published over 40 peer-reviewedjournal and conference papers. His work in continental-scale high-resolution floodmapping and political redistricting have received broad media coverage by the National Science Foundation (NSF), HPC Wire, Top 500, WIRED, Communications of the ACM, and Nature.
报告摘要:
Spatial optimization (SO) is an important and prolific field of interdisciplinary research. Spatial optimization methods seek optimal allocation or arrangement of spatial units under spatial constraints such as distance, adjacency, contiguity, partition,etc.As spatial granularity becomes finer and problem formulations incorporate increasingly complex compositions of spatial information, the performance of spatial optimization solvers becomes more imperative. My research focuses on scalable spatial optimizationmethodswithin the evolutionary algorithm (EA) framework. The computational scalability challenge in EA is addressed by developing a parallel EA library that eliminates the costly global synchronization in massively parallel computing environment and scalesto 131,072processors. A spatially explicit EA framework that couples graph representations of spatial constraints with intelligent guided search heuristics such as path relinking and ejection chain is proposed to effectively explore SO decision space. Thisframeworkis employed to solve large political redistricting problems and create billions of feasible districting plans that adhere to Supreme Court mandates.
邀请人：吴产乐教授，吴黎兵教授
