Process-Monitoring-for-Quality — Big Models

Abstract Process Monitoring for Quality (PMQ) is a big data-driven quality philosophy aimed at defect detection (through binary classification) and empirical knowledge discovery. It was originally developed to solve a complex manufacturing quality problem. It is founded on Big Models , a predictive modeling paradigm based on machine learning, statistics and optimization, that includes a learning aspect that requires many models to be developed to find the final model . When dealing with big data, the data structure is not known in advance; therefore, there is no a priori distinction between learning algorithms, and a plethora of options to choose from. The learning scheme of Big Models is described, which is based on several well known learning algorithms with the capacity to effectively solve a wide spectrum of binary classification problems. The main challenges of manufacturing pattern recognition problems are discussed and addressed to provide a strong foundation to the Big Models learning paradigm. Finally, two defect detection case studies are presented with highly unbalanced data derived from real manufacturing systems to validate the proposal.