Use of Decision-Tree Induction for Process Optimization and Knowledge Refinement of an Industrial Process

Development of expert systems involves knowledge acquisition which can be supported by applying machine learning techniques. This paper presents the basic idea of using decision-tree induction in process optimization and development of the domain model of electrochemical machining (ECM). It further discusses how decision-tree induction is used to build and refine the knowledge base of the process. The idea of developing an intelligent supervisory system with a learning component (IMAFO, Intelligent MAnufacturing FOreman) that is already implemented is briefly introduced. The results of applying IMAFO for analyzing data form the ECM process are presented. How the domain model of the process (electrochemical machining) is built from the initial known information and how the results of decision-tree induction can be used to optimize the model of the process and further refine the knowledge base are shown. Two examples are given to demonstrate how new rules (to be included in the knowledge base of an expert system) are generated from the rules induced by IMAFO. The procedure to refine these types of rules is also explained.Le d\ue9veloppement de syst\ue8mes experts facilite l'acquisition de connaissances gr\ue2ce \ue0 des techniques d'apprentissage machine. Cet article pr\ue9sente l'id\ue9e de base qui consiste \ue0 utiliser l'induction d'arbre de d\ue9cision dans l'optimisation du processus et le d\ue9veloppement du mod\ue8le de domaine de l'usinage \ue9lectrolytique (ECM ou electrochemical machining). On examine plus en d\ue9tail comment l'induction d'arbre de d\ue9cision sert \ue0 b\ue2tir et \ue0 affiner la base de connaissances du processus. On explique bri\ue8vement l'id\ue9e de d\ue9velopper un syst\ue8me de supervision intelligent avec une composante d'apprentissage (IMAFO ou Intelligent MAnufacturing Foreman) qui est d\ue9j\ue0 implant\ue9e. Les r\ue9sultats de l'application de IMAFO \ue0 l'analyse des donn\ue9es forment le processus ECM. On montre comment le mod\ue8le de domaine du processus (usinage \ue9lectrolytique) est b\ue2ti \ue0 partir de l'information connue au d\ue9part et comment les r\ue9sultats de l'induction d'arbre de d\ue9cision permettent d'optimiser le mod\ue8le du processus et d'affiner encore plus la base de connaissances. Deux exemples illustrent comment les nouvelles r\ue8gles (\ue0 inclure dans la base de connaissances d'un syst\ue8me expert) sont g\ue9n\ue9r\ue9es d'apr\ue8s les r\ue8gles induites par IMAFO. La proc\ue9dure d'affinement de ce type de r\ue8gles est \ue9galement expliqu\ue9e.NRC publication: Ye