Cost-sensitive learning classification strategy for predicting product failures

摘要: Abstract In the current era of Industry 4.0, sensor data used in connection with machine learning algorithms can help manufacturing industries to reduce costs and predict failures advance. This paper addresses a binary classification problem found engineering, which focuses on how ensure product quality delivery at same time production costs. The aim behind this is number faulty products, case extremely low. As result characteristic, reduced an imbalanced problem. authors contribute research three important ways. First, industrial application coming from electronic industry presented detail, along its modelling challenges. Second, modified cost-sensitive strategy based combination Voronoi diagrams genetic algorithm applied tackle compared several base classifiers. results obtained are promising for specific application. Third, order evaluate flexibility strategy, demonstrate wide range applicability, 25 real-world sets selected KEEL repository different imbalance ratios features. implemented without predefined cost, classifiers as those