Description
This workshop will introduce the essential machine learning algorithms and software tools for graduate students, experienced researchers and engineers working in the industry. Elementary knowledge of probability and statistics is required to attend this workshop. The workshop will also feature at least two confirmed guest speakers with decades of experience in data analytics. Prof. Sirish L. Shah of University of Alberta and Prof. Richard D. Braatz of Massachusetts Institute of Technology have both agreed to speak during the workshop. We have also invited an industrial guest speaker.
Workshop Overview
We are currently at the cusp of what is considered the fourth industrial revolution. This revolution is driven by the ubiquitous cyber-physical systems, algorithmic developments in artificial intelligence, gargantuan computing power, inexpensive memory and the gigantic volumes of data that are being collected. The process industries are in possession of treasure troves of heterogenous data that is gravely under utilized. The competitive global environment, and the ever increasing demands on energy, environment and quality are subjecting these industries to a high level of economic pressure. The incredible volumes of data that they already possess are poised to provide a level of automation and efficiency never seen before and thus alleviate the economic and competitive pressures.
Process industries have been using data analytics in various forms for more than three decades. In particular, statistical techniques such as principal component analysis (PCA), partial least squares (PLS), canonical variate analysis (CVA) and time series methods for modeling such as maximum likelihood estimation, prediction error methods have been extensively applied on industrial data. The recent developments in machine learning and artificial intelligence provide a new opening for using process data on large scale problems. However, in order to successfully apply machine learning methods to process data, researchers require not only a high level understanding of the algorithms but also strong programming knowledge in packages such as Python, TensorFlow, Keras and Jupyter.
Guest Speakers
Dr. Richard Braatz
Dr. Richard D. Braatz is the Edwin R. Gilliland Professor of Chemical Engineering at the Massachusetts Institute of Technology (MIT) where he does research in applied mathematics and control theory and their application to chemical and biological systems. He received an MS and PhD from the California Institute of Technology and was the Millennium Chair and Professor at the University of Illinois at Urbana-Champaign and a Visiting Scholar at Harvard University before moving to MIT. He has consulted or collaborated with more than 20 companies including IBM, United Technologies Corporation, Novartis, and Abbott Laboratories. Honors include the Donald P. Eckman Award from the American Automatic Control Council, the Curtis W. McGraw Research Award from the Engineering Research Council, and the AIChE Computing in Chemical Engineering Award. He is a Fellow of the Institute of Electrical and Electronics Engineers, International Federation of Automatic Control, and the American Association for the Advancement of Science. For more information, see Dr. Braatz's page.
Dr. Sirish Shah
Dr. Sirish L. Shah has been with the University of Alberta since 1978, where he held the NSERC-Matrikon-Suncor-iCORE Senior Industrial Research Chair in Computer Process Control from 2000 to 2012. He is the recipient of the Albright & Wilson Americas Award in 1989, the Killam Professor in 2003, the D.G. Fisher Award for significant contributions in the field of systems and control, the ASTECH award in 2011 and the 2015-IEEE Transition to Practice Award. He has held visiting appointments at Oxford University and Balliol College as a SERC fellow, Kumamoto University (Japan) as a senior research fellow of the Japan Society for the Promotion of Science (JSPS), the University of Newcastle, Australia, IIT-Madras India and the National University of Singapore. The main areas of his current research are process and performance monitoring, analysis and rationalization of alarm systems. He has co-authored three books, the first titled, Performance Assessment of Control Loops: Theory and Applications, a second titled ‘Diagnosis of Process Nonlinearities and Valve Stiction: Data Driven Approaches”, and a more recent monograph on “Capturing connectivity and causality in complex industrial processes”. He is emeritus professor at the University of Alberta, a fellow of the Canadian Academy of Engineering and the Chemical Institute of Canada. For more information, visit Sirish Shah's page.
Organizers
Dr. Bhushan Gopaluni
Dr. Bhushan Gopaluni is a professor in the department of chemical and biological engineering and an Associate Dean for Education and Professional Development in the faculty of Applied Science at the University of British Columbia. He is also an associate faculty in the Institute of Applied Mathematics, the Institute for Computing, Information and Cognitive Systems, Pulp and Paper Center and the Clean Energy Research Center. He is currently an associate editor for Journal of Process Control, The Journal of Franklin Institute, guest editor for Process Control Special Series in the Canadian Journal of Chemical Engineering. He received a Ph.D. from the University of Alberta in 2003 and a Bachelor of Technology from the Indian Institute of Technology, Madras in 1997 both in the filed of chemical engineering. From 2003 to 2005 he worked as an engineering consultant at Matrikon Inc. (now Honeywell Process Solutions) during which he had designed and commissioned multivariable controllers in British Columbia’s pulp and paper industry, and had implemented numerous controller performance monitoring projects in the Oil & Gas and other chemical industries. He is one of the leading experts on data analytics for process industry and has authored over 110 refereed articles in reputed international Journals and conferences. His publications have been recognized through best paper awards and keynote presentations. He is also the recipient of the prestigious Killam Teaching Prize and the Dean’s service medal from the University of British Columbia. For more information, visit the DAIS page.
Lee Rippon
Lee Rippon is a PhD student studying Chemical and Biologial Engineering (CHBE) at UBC. He also holds BASc and MASc degrees from UBC in CHBE where his research experience includes applications of compressive sensing, adaptive control, system identification and process monitoring on sheet and film processes. His current research interests include applying statistical machine learning techniques to historical process data to perform fault detection, isolation, and diagnosis in a kraft process. For more information, visit the DAIS page.
Yiting Tsai
Yiting Tsai has finished both his BASc and MASc degrees at CHBE. His interests are process control and statistical modeelling of time-series data. His current PhD research focuses on the application of Machine Learning techniques to design smart controllers, which identify and predict process faults ahead of time and apply appropriate control actions to prevent such faults. For more information, visit the DAIS page
Dr. Aditya Tulsyan
Dr. Aditya Tulsyan is currently a Senior Engineer at Amgen. Prior to joining Amgen in 2016, Aditya was a Postdoctoral Associate in the Process Systems Engineering Laboratory at the Massachusetts Institute of Technology. He received his Ph.D. in Computer Process Control from the University of Alberta, Canada in 2013. He has held research positions at the National University of Singapore, University of British Columbia and the Indian Institute of Technology, Kharagpur. His research interests are in systems engineering, statistical machine learning, signal processing and Bayesian inference. For more information, visit Dr. Tulsyan's page.
Course Plan
Starting with an elementary introduction to statistics and probability, we will develop various regression, classification, dimensionlity reduction and advanced learning algorithms that are of interest to engineers. In addition, various widely-used machine learning software packages will be introduced. Registrants will solve exercises and receive take-away software code to implement these algorithms. The following is a general outline of the course:
- Basics of probability and statistics, underfitting, overfitting and bias-variance tradeoff
- Classification Algorithms
- Support Vector Machines
- Naive Bayes Classifier
- Regression Algorithms
- Linear Least Squares
- Kernel Regression
-
Dimensionality Reduction Algorithms
- Principal Component Analysis (PCA)
- Partial Least Squares (PLS)
- Isometric Mapping (ISOMAP)
- Advanced Learning Algorithms
- Deep Learning
- Recurrent Neural Networks
- Gaussian Processes
- Applications in the Process Industry
Learning Outcomes
By the end of this workshop, registrants will be able to:
- identify and solve classification, regression and dimensionality reduction problems
- work with softwares such as Python, TensorFlow, and Keras
Schedule
8:30AM - 9:30AM
Richard Braatz
Big Data Analytics
9:30AM - 10:30AM
Bhushan Gopaluni
Classification
10:30AM - 11:30AM
Yiting Tsai
Regression
11:30AM - 12:30PM
Bhushan Gopaluni, Yiting Tsai
Dimensionality Manipulation
12:30PM - 1:30PM
Lunch
1:30PM - 2:30PM
Sirish Shah
Alarm Management
2:30PM - 5:30PM
Bhushan Gopaluni, Aditya Tulsyan, Yiting Tsai
Advanced Learning Algorithms
Software
This workshop will be using Python code in Jupyter notebooks. If you would like to follow along with the code during the workshop, it is recommended that you install the required software.
For a guide to installation, follow the TensorFlow pages provided below. The TensorFlow page contains instructions on setting up the environment properly.