Accession Number : ADA294087

Title :   A Multistrategy Learning Scheme for Assimilating Advice in Embedded Agents

Corporate Author : NAVAL RESEARCH LAB WASHINGTON DC

Personal Author(s) : Gordon, Diana F.

PDF Url : ADA294087

Report Date : 1993

Pagination or Media Count : 16

Abstract : The problem of designing and refining task-level strategies in an embedded multiagent setting is an important unsolved question. To address this problem, we have developed a multistrategy system that combines two learning methods: operationalization of high-level advice provided by a human and incremental refinement by a genetic algorithm. The first method generates seed rules for finer-grained refinements by the genetic algorithm. Our multistrategy learning system is evaluated on two complex simulated domains as well as with a Nomad 200 robot. (AN)

Descriptors :   *RULE BASED SYSTEMS, *KNOWLEDGE BASED SYSTEMS, *LEARNING, ALGORITHMS, COMPUTERIZED SIMULATION, SCENARIOS, NEURAL NETS, OPTIMIZATION, PARAMETRIC ANALYSIS, STRATEGY, COMPARISON, REASONING, ROBOTS, LEARNING MACHINES, PROBLEM SOLVING, CONVERGENCE, SYSTEMS ANALYSIS, CONTROL THEORY, REFINING, COMPILERS, HIGH LEVEL LANGUAGES, GAME THEORY.

Subject Categories : Cybernetics

Distribution Statement : APPROVED FOR PUBLIC RELEASE