Methods of Validating Expert Systems
Generate an AI Snapshot to get a quick, structured summary of this paper.
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
Validating expert systems is critical because without appropriate validation, the system may make costly errors and be unable to determine the quality of its decisions.
Abstract
Validating expert systems is critical. Without appropriate validation, the system may make costly errors. There are a number of methods of validating expert systems. They can be validated by analyzing and testing the knowledge base and inference engine individually. However, any analysis of the system components is limited to looking at pieces of the system and not at how they work together. The system as a whole can be analyzed by viewing it as a black box and determining the quality of its decisions or by opening up the black box and determining why it made particular decisions.