Modulbeschreibung
Machine Learning and Data Science (EEU)
Kürzel:
M_MLDS
Durchführungszeitraum:
HS/24
ECTS-Credits:
4
Lernziele:
The students learn
- the fundamentals of Machine Learning (ML) for data analysis
- data analysis and data processing in Python
- the application of Machine Learning problems in energy and environmental technology.
- to view the learning problem as an optimization process, to solve it, and to evaluate the quality of the results
- to solve technical problems in mixed teams using ML-based data analysis
Verantwortliche Person:
Prof. Dr. Nordborg Henrik
Standort (angeboten):
Rapperswil-Jona
Vorausgesetzte Module:
Skriptablage:
Modultyp:
Wahlpflicht-Modul für Erneuerbare Energien und Umwelttechnik STD_21(Empfohlenes Semester: 5)
Wahlpflicht-Modul für Erneuerbare Energien und Umwelttechnik STD_24(Empfohlenes Semester: 5)
ECTS-Credits pro Kategorie
Spezialkategorie: Grundlagen EEU, Vertiefung ET, Vertiefung UT, Mathematik, Naturwissenschaften / 4 Credits
Vertiefungsmodule EEU / 4 Credits
Modulbewertung
Bewertungsart:
Note von 1 - 6
Leistungsbewertung
Während der Prüfungssession:
Mündliche Prüfung, 30 Minuten
Während des Semesters:
Seminararbeit
Bewertungsart:
Note von 1 - 6
Gewichtung:
Seminararbeit (50%), mündliche Prüfung (50%)
Bemerkungen:
Kurse in diesem Modul
Machine Learning and Data Science
Kürzel:
MLDS
Semester:
1
Lernziele:
- know the different approaches to AI-based data analysis (e.g. supervised vs. unsupervised learning)
- process and analyze data with Python
- formulate adequate models for problems from EU and UT and train them on specific data sets
- understand the training process as an optimization problem, solve it and evaluate the quality of the result qualitatively and quantitatively
- solve technical problems in mixed teams using AI-based data analysis
Plan und Lerninhalt:
-
(Note: not in chronological order)
- Introduction to data analysis with Python
- Review of probability (joint, marginal, and conditional probability; Baye's rule; probability density, ...)
- Review of linear algebra (matrices and vectors, eigen-values and -vectors, factorization, ...)
- Fundamentals of ML-base data analysis
- Formulation of the learning problem
- Types of learning according to data availability and type
- supervised learning: regression and classification
- unsupervised learning: clustering and dimensionality reduction
- reinforcement learning
- semi-supervised learning
- causal learning
- Methods
- Constrained and regularized optimization
- Linear and non-linear regression
- Gaussian processes and support vector machines
- Neural networks
- Kalman filter and reservoir computing (iterative learning)
- Q-learning
- K-means
- PCA, ICA, and extensions (e.g. kernel PCA)
- Applications in the fields of renewables and environmental technology
- Other relevant topics
- The relation between ML and AI
- Interpolation, Extrapolation
- Filtering, smoothing, and forecasting
- Model selection Overfiting and classes of functions
Unterrichtssprache:
Englisch
Bibliographie:
- Abu-Mostafa, Y. S., Magdon-Ismail, M., & Lin, H.-T. (2012). Learning From Data. AMLBook.
- Bishop, C. M. (2016). Pattern Recognition and Machine Learning (Softcover reprint of the original 1st edition 2006 (corrected at 8th printing 2009)). Springer New York.
- Rasmussen, C. E., & Williams, C. K. I. (2006). Gaussian Processes for Machine Learning. MIT Press. http://www.gaussianprocess.org/gpml/chapters
- Brunton, S. L., & Kutz, J. N. (2022). Data-Driven Science and Engineering: Machine Learning, Dynamical Systems, and Control (2nd ed.). Cambridge University Press. https://doi.org/10.1017/9781009089517
- Peters, J., Janzing, D., & Schölkopf, B. (2017). Elements of causal inference: Foundations and learning algorithms. MIT Press.
- Särkkä, S. (2013). Bayesian Filtering and Smoothing. Cambridge University Press. https://users.aalto.fi/~ssarkka/#books
Kursart:
Durchführung gemäss Stundenplan
Vorlesung mit 2 Lektionen pro Woche
- Max. Teilnehmer: 72
- Harte Grenze: ja
Uebung mit 2 Lektionen pro Woche
- Max. Teilnehmer: 18
- Harte Grenze: ja