Application of Adaptive Decision Aiding Systems to Computer-Assisted Instruction
Abstract
A minicomputer-based Computerized Diagnostic and Decision Training (CDDT) system, combining principles of artificial intelligence, decision theory, and adaptive computer-assisted instruction, is described. Training focusses on decision-making in electronic troubleshooting. The CDDT System incorporates an adaptive computer program which learns the student's diagnostic and decision value structure using a trainable network technique of pattern classification, compares this structure to that of an expert, and adapts the instructional sequence to eliminate discrepancies through the use of heuristic algorithms. An expected value model of decision-making is the basis of the student and instructor models which, with the task simulator and adaptive instruction, form the core of the CDDT system. The instructor model also generates suggested actions in response to student requests for assistance. The student's task is to troubleshoot a complex circuit by making test measurements, replacing the malfunctioning part, and making verification measurements. The student values of interest are those for information gained through the measurements, and for the replacement of circuit modules.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1977
- Accession Number
- ADA055657
Entities
People
- Amos Freedy
- Denis D. Purcell
- Donald M. May
- Luigi F. Lucaccini
- William H. Crooks