GENERAL BAYESIAN MODELS FOR MULTIATTRIBUTE ACCEPTANCE SAMPLING (QUALITY CONTROL)
Abstract
Classical acceptance sampling systems, such as MIL-STD-105D and Dodge & Romig Tables, have been the most widely used quantitative tools in quality control. Recent developments in the Bayesian approach emphasize an explicit consideration of economic consequences associated with decisions to accept or reject inspection lots. In contrast, the statistical criteria used in the classical approach are often arbitrary. Researchers in both areas have concentrated on the selection of sampling plans for single attribute inspection rather than a simultaneous consideration of all attributes subject to control. In this research, general Bayesian multiattribute models which accommodate various dispositions of rejected lots are developed. Two inspection procedures, (1) full inspection and (2) stepwise inspection, are used to evaluate the model. The models are investigated both analytically and empirically. An iterative solution procedure is also developed and shown to be very effective in obtaining an optimal multiattribute sampling plan. Several possible heuristic sequencing rules for the stepwise inspection procedure are evaluated by numerical experiments.
Degree
Ph.D.
Subject Area
Business community
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.