Use of Machine Learning Methods in Decision Making to Ensure Quality in Instrument Manufacturing

Abstract

Quality assurance is an important and various activity that has a significant impact on the competitiveness of both the organization's products and its management system as a whole. This activity is not a value-creating main process and is managed as a cost center, which creates objective difficulties in increasing its efficiency. Despite the large number of publications in the field of quality assurance, the topic of decision support in this area has been poorly researched and is considered mainly for high-level managers. Conceptually, the problem of increasing the efficiency of decisions is considered in a number of publications of the author. In this paper, this approach is considered for quality assurance specialists at the level of the executer or expert and is used for the first time in this formulation of the problem. The aim of the study is, in accordance with the previously proposed methodology, the formation of a situational-ontological model for solving a specific practical problem of identifying the cause of a defect for an instrument-making enterprise using the machine learning method (k-nearest neighbors). The results of the study allowed us to conclude that it is possible to use the situational ontological methodology to create a decision support system for solving problems in a typical situation of finding the cause of a defect based on the k-nearest neighbors method for a quality engineer of an instrument-making enterprise.