Machine Learning in Chemical Engineering
Motivation
The application of Industry 4.0 concepts made it possible for plant operators to better manage productivity, energy efficiency, and safety in production. Modern plants are now highly automated, inter-connected and extensively equipped with sensors. For a further improvement, plant operators are exploiting a huge amount of production data generated during operation. Data-driven modeling methods or so-called machine learning are used for this purpose.
Content
The course deals with all aspects along the entire life cycle to solve a problem using machine learning methods. Contents of the course are the selected methods, models and tools for data-driven modeling of industrial processes. The focus is on the problem-oriented application of machine learning approaches to solve typical tasks such as regression, classification, clustering and time series analysis. The distinguishing feature of this course is the consideration of typical features in process engineering.
Main content:
- Workflow for conducting ML projects
- Data acquisition, exploration and visualization
- Data preprocessing (cleaning, feature engineering, PCA)
- Traditional ML models, target functions and metrics for regression, classification and clustering problems
- Numerical methods for parameter estimation and regularization approaches
- Feed-forward neural networks, activation functions, modern architectures, transfer learning, software frameworks
- Basics of time series analysis, time series models
In addition to the lectures, there are excersise, where the lecture contents are deepened with the help of practice-oriented and industry-related problems. The core of the exercises is the practical use of ML-typical software packages like Jupyter Notebook, Pandas, Scikit-Learn and PyTorch.
Two self-paced assignment round off the course.
The course is taught in English.
Scope
(2 2 0 SWS)
Lecture Documents
More information can be found Machine Learning in Chemical Engineering at OPAL.
Recommended Prerequisites
- Basics of Linear Algebra
- Basics of Statistics