Using Discovery-Based Learning to Prove the Behavior of an Autonomous Agent
Abstract
Computer-generated autonomous agents in simulation often behave predictably and unrealistically. These characteristics make them easy to spot and exploit by human participants in the simulation, when we would prefer the behavior of the agent to be indistinguishable from human behavior. An improvement in behavior might be possible by enlarging the library of responses, giving the agent a richer assortment of tactics to employ during a combat scenario. Machine learning offers an exciting alternative to constructing additional responses by hand by instead allowing the system to improve its own performance with experience. This thesis presents NOSTRUM, a discovery-based learning (DBL) system designed to work in tandem with the MAXIM air combat simulator. Through a process of repeated experimentation modeled after the scientific method, NOSTRUM was able to discover many responses that were more appropriate than the single mode of agent control implemented in the original MAXIM program. NOSTRUM often found responses that dramatically improved the offensive position of the agent, and it sometimes placed the agent in position for an extended shot on the target when one was not available before.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1993
- Accession Number
- ADA274131
Entities
People
- David N. Mezera
Organizations
- Air Force Institute of Technology