Jiming Liu (Computer Science, HKBU) 

Title: Autonomy Oriented Computing (AOC): A New Paradigm in Data Mining 
and Modeling

Abstract: A multi-agent system consists of computational entities that 
readily demonstrate the capabilities as well as properties of 
distribution, autonomy, reactiveness, learning adaptation, and 
self-organization in response to their local inputs or environments.  In 
this talk, we will introduce a new direction of research, called Autonomy
Oriented Computing (AOC), that, to a certain extent, departs from the 
multi-agent systems research. Specifically speaking, AOC explores novel 
computing paradigms in using multi-agent systems to characterize complex 
systems behaviors and to solve hard computational problems. In this talk, 
we will examine how AOC works in data mining and modeling, by means of 
discussing some representative issues and examples; namely, foraging agent 
based regularity characterization, autonomous pattern search in constrained 
domains, and autonomous on-line mining in unstructured domains.