Machine Learning 1
General Information
- Machine Learning 1 is a 9 LP (9 ECTS) credits module.
- Machine Learning 1-X is a 12 LP (12 ECTS) credits module.
Lectures period |
18.10.2021 - 19.02.2022 |
Q&As |
Tuesdays, 14:15 - 16:00 (online) |
Exercises |
Wednesdays, 16:15 - 18:00 (online) |
Lecture |
Thursdays, 14:15 - 16:00 (online) |
Trainers |
Klaus-Robert Müller |
Grégoire Montavon |
|
Contact |
|
ISIS |
|
Language |
English |
Frequently asked questions (FAQ):
I don't have a TU account yet, how do I connect to the course? On the ISIS course website, you can login with Guest Access (password: Ahc8tiet).
How to register for the course? There is no pre-registration. Just come to the first lecture.
What are the prerequisites? There are no formal prerequisites. Good knowledge of linear algebra, calculus, probability theory, and programming, as well as some machine learning basics are however recommended.
I am from a different university, can I take this course? If you are not a student at TU and want to earn credit, you have to solicit ''Nebenhörerschaft''.
Prerequisites
The following are recommended prerequisites which are helpful but not necessary for taking the course:
- Good knowledge in linear algebra and calculus, as presented in the respective modules (German: Lineare Algebra, Analysis)
- Good knowledge in probability theory, as presented in the module stochastics (German: Elementare Stochastik)
- Good programming knowledge, programming in Python
- Machine learning basics (e.g. classification).
As a thematic preparation, it is possible to visit the Python course which is also accreditable as optional compulsory course parts.
Topics
The scheduled topics are:
- Bayesian ML
- Bayes Decision Theory
- Maximum Likelihood Estimation and Bayes Parameter Estimation
- Analyses
- Principal Component Analysis
- Linear Discriminant Analysis
- Machine Learning Theory
- Model Selection and Bias/Variance Tradeoff
- VC Dimension and Kernels
- Classification and Regression
- Support Vector Machines
Decision Trees & Random Forests
- Boosting
- Kernel Ridge Regression
- Neural Networks and Backpropagation
- Latent Variable Models
- k-means Clustering
- Expectation Maximization
- Restricted Boltzmann Machines