Process Data Analytics and Machine Learning

S. Joe Qin, Leo H. Chiang, and Richard D. Braatz

This workshop will cover data analytics and machine learning methods that are applied to chemical processes. The workshop consists of three modules, each of which includes presentations, discussions, and case studies.

As of July 26, there are no more workshop slots available.

Agenda

1 - Introduction

• 1.1 Examples of typical data analytics applications
• 1.2 Unsupervised, supervised, and partially supervised learning (including clustering)
• 1.3 Least squares including sparse methods
• 1.4 Feature engineering (including for categorical data)
• 1.5 Kernel methods for nonlinear analytics
• 1.6 Neural networks and deep learning

2 - Latent Variable Methods and Application Case Studies

• 2.1 Principal component analysis
• 2.2 Partial least squares
• 2.3 Canonical correlation analysis
• 2.4 Dynamic principal component analysis and canonical variate analysis
• 2.5 Linear discriminant analysis and support vector machines
• 2.6 Process monitoring, diagnosis, and troubleshooting

3 - Industrial Experience and Tips, Interactive Discussions

• 3.1 Visualization
• 3.2 Outlier detection and data preprocessing
• 3.3 Method selection
• 3.4 How good is good enough? Industrial tips and tricks of the trade
• 3.5 Industrial case studies by guest speaker Dr. Ivan Castillo (Dow)

This workshop is scheduled before the FOPAM conference. The workshop will be on Monday, August 5th from 1:30 pm to 5:00 pm and on Tuesday, August 6th from 8:30 am to 5:00 pm.

The fee for the workshop is $500 and you can sign up to attend the workshop on the registration form.