Agent-Based Computing

Posted 1 year ago by Munehiro Fukuda

School(s) : STEM
Primary PI Name : Munehiro Fukuda
Email : mfukuda@uw.edu
Phone : 425-352-3459
Project/Faculty Website : http://faculty.washington.edu/mfukuda/
Research Location : UW Bothell
Project Goals : Multi-agent simulations: model a large number of social or biological agents, (each with an independent code and data) and simulate their emergent collective behavior that may be difficult only with mathematical and macroscopic approaches. For instance, FluTE: simulates a nationwide influenza epidemic by generating 280 million agents, based on 2000 US Census. Agent-based data analysis: injects agents into big data for carrying out intelligent and pinpoint data search or analysis, which we believe is much more efficient than conventional data-analyzing software such as MapReduce or Spark that scans all data items greedily. MASS: Multi-Agent Spatial Simulation is a computer language library to develop and execute agent-based simulation or agent-based data analysis with multiple computers. Distributed Systems Laboratory (DSL) at the CSS division is developing MASS in Java, C++, and CUDA. This project seeks (1) to assist the MASS development team in developing benchmark test programs, each modeling core logics of a different multi-agent simulation (in biology, business, or social sciences) or data analysis (in GIS or machine learning), or (2) to enhance (a.k.a. hack) the MASS library itself: functionality enhancement, speed-up, and GPU programming.
Student Qualifications : Fluent with Java (CSS162 completed) or C++ programming (CSS342 completed), commit to project for two quarters, must register for CSS497 or CSS499.
Student Outcomes : Attend weekly research group meetings. Present a 30-minute research work in DSL each quarter. Demonstrate software deliverables in DSL at the end of quarter. Document the research outcome as a term paper.
Time Frame : Minimum 2 Quarters.

  • School(s) : STEM
  • Primary PI Name : Munehiro Fukuda
  • Interested? Contact Faculty Researcher by Email : mfukuda@uw.edu
  • Phone : 425-352-3459
  • Project/Faculty Website : http://faculty.washington.edu/mfukuda/
  • Research Location : UW Bothell
  • Project Goals : Multi-agent simulations: model a large number of social or biological agents, (each with an independent code and data) and simulate their emergent collective behavior that may be difficult only with mathematical and macroscopic approaches. For instance, FluTE: simulates a nationwide influenza epidemic by generating 280 million agents, based on 2000 US Census. Agent-based data analysis: injects agents into big data for carrying out intelligent and pinpoint data search or analysis, which we believe is much more efficient than conventional data-analyzing software such as MapReduce or Spark that scans all data items greedily. MASS: Multi-Agent Spatial Simulation is a computer language library to develop and execute agent-based simulation or agent-based data analysis with multiple computers. Distributed Systems Laboratory (DSL) at the CSS division is developing MASS in Java, C++, and CUDA. This project seeks (1) to assist the MASS development team in developing benchmark test programs, each modeling core logics of a different multi-agent simulation (in biology, business, or social sciences) or data analysis (in GIS or machine learning), or (2) to enhance (a.k.a. hack) the MASS library itself: functionality enhancement, speed-up, and GPU programming.
  • Student Qualifications : Fluent with Java (CSS162 completed) or C++ programming (CSS342 completed), commit to project for two quarters, must register for CSS497 or CSS499.
  • Student Outcomes : Attend weekly research group meetings. Present a 30-minute research work in DSL each quarter. Demonstrate software deliverables in DSL at the end of quarter. Document the research outcome as a term paper.
  • Number of Student Positions Available : 2
  • Time Frame : Minimum 2 Quarters.