Decision Trees And Quality Control Decisions

Abstract

Bayesian decision tree analysis has been widely used as a basis for quality control decision making.  Recently, the traditional decision tree analysis has been criticized for requiring a lot of calculations and, therefore, being inefficient.  This paper presents a simplified and efficient decision tree analysis for quality control decision making that improves the efficiency of the traditional decision analysis by reducing substantially the number of calculations required to solve decision problems.  For some decision problems, the proposed analysis reduces the number of calculations required to solve decision problems by more than 75%.

Some researchers provided modified decision trees (Game trees and Scenario trees) that attempt to preserve the advantages of the traditional trees while improving their efficiency.  However, these other modified decision trees may not be as efficient as the traditional analysis because they do not allow for the use of the coalescence procedure in the case of symmetrical decision problems.