Lehrinhalte
The analysis and interpretation of data becomes ever more important, also for engineers. Digitalization and Smart Grids are terms to describe a host of novel data-based services in the field of generation, distribution, consumption, and marketing of (renewable) energy. The lecture presents the recent developments and their underlying machine learning methods.
For a start we describe the different problem settings of machine learning methods, review recent developments in the field, and evaluate the impact of machine learning on the energy sector. After such an introductory overview, we review the basics of linear algebra and numerical optimization. We then introduce supervised learning problems and study different model classes to solve such problems (linear models, trees, random forests, nearest neighbor, kernel methods, deep learning). We then turn to a probabilistic view and study unsupervised learning problems. Finally, we give an introduction to probabilistic graphical models. Throughout the semester we discuss exemplary applications of machine learning in the energy domain (e.g. renewable forecasting, predictive maintenance, state estimation, probabilistic load flow).
Practical exercises with Python deepen the understanding and support students actively usable skills.
Literature
[list]
[*]K.P. Murphy: Machine Learning. A Probabilistic Perspective.
[*]C.M. Bishop: Pattern Recognition & Machine Learning
[*]J. Friedman, T. Hastie, R. Tibshirani: The elements of statistical learning
[*]D. Koller, N. Friedmann: Probabilistic Graphical Models. Principles and Techniques
[/list]
Voraussetzungen
[list]
[*]Good knowledge of linear algebra required
[*]Basic knowledge of statistics and numerical optimization will be helpful
[*]Using Python for programming the practical examples should pose no difficulty
[/list]
The analysis and interpretation of data becomes ever more important, also for engineers. Digitalization and Smart Grids are terms to describe a host of novel data-based services in the field of generation, distribution, consumption, and marketing of (renewable) energy. The lecture presents the recent developments and their underlying machine learning methods.
For a start we describe the different problem settings of machine learning methods, review recent developments in the field, and evaluate the impact of machine learning on the energy sector. After such an introductory overview, we review the basics of linear algebra and numerical optimization. We then introduce supervised learning problems and study different model classes to solve such problems (linear models, trees, random forests, nearest neighbor, kernel methods, deep learning). We then turn to a probabilistic view and study unsupervised learning problems. Finally, we give an introduction to probabilistic graphical models. Throughout the semester we discuss exemplary applications of machine learning in the energy domain (e.g. renewable forecasting, predictive maintenance, state estimation, probabilistic load flow).
Practical exercises with Python deepen the understanding and support students actively usable skills.
Literature
[list]
[*]K.P. Murphy: Machine Learning. A Probabilistic Perspective.
[*]C.M. Bishop: Pattern Recognition & Machine Learning
[*]J. Friedman, T. Hastie, R. Tibshirani: The elements of statistical learning
[*]D. Koller, N. Friedmann: Probabilistic Graphical Models. Principles and Techniques
[/list]
Voraussetzungen
[list]
[*]Good knowledge of linear algebra required
[*]Basic knowledge of statistics and numerical optimization will be helpful
[*]Using Python for programming the practical examples should pose no difficulty
[/list]
- Lehrende: Anonym
- Lehrende: Allan Santos
- Lehrende: Florian Steinke
Semester: WT 2021/22