Information System Lectures  

04.07.16 00:00

Machine Learning Evaluation for Unary Classification in Fault Diagnosis, Prof. Dr. Adamo Santana

Montag, 4. Juli 2016 - 16:00 Uhr - Seminarraum 5 (I 127)

Professor Adamo Santana

Abstract

Early fault detection of equipment is important for any industry, but as more reliable products, with longer working life, are produced, historical data on fault events becomes scarce; moreover, the available data is usually not labelled, another aspect that is essential for the proper training of classification algorithms. Following this premise, to classify events as in-control of faulty, samples or patterns for both types of events are needed, however, it is often the case in practical industrial applications where only the in-control type of data is available. While the use of multivariate statistical process control (MSPC) methods and metrics are the often applied strategies for monitoring in such cases, the direct application of machine learning algorithms, supervised and unsupervised, also prove as viable alternatives in the task of identifying unusual behavior in real world data.


Short biography

Adamo Santana is a Professor at the Institute of Technology of the Federal University of Pará (UFPA), in Brazil, where he coordinates the Laboratory of Computational Intelligence and Operational Research, developing researches in the areas of machine learning and data mining for decision support. He received the bachelor's degree in Computer Science from the University of the Amazon in 2002; the master's (2005) and Ph.D. (2008) in Applied Computing for Electric Engineering from UFPA, with collaborative period in the University College Cork. Currently, he is a Guest Professor at the School of Interdisciplinary Mathematical Sciences of Meiji University, in Japan, acting on the subjects of optimization for power systems.



letzte Änderung: 10.11.2016