Machine Learning for Robotics and Computer Vision (IN3200) (2h + 2h, 5ECTS)
WS 2017, TU München
You can use our library for the programming exercises: mlcv-tutorial
Beginning from Monday, 13.11.2017, the tutorial will be taking place in Room 00.08.059.
There is no tutorial on Monday, 20.11.2017.
Some students noted that their exam registration status in TUMonline is: "registered (preliminary registration)" with an exclamation mark in yellow circle. As far as we know that does not affect you. You can come to the exam.
No cheatsheets, calculators or other assistances are allowed.
There is NO repeat exam. The course is offered again in the next semester.
In this lecture, the students will be introduced into the most frequently used machine learning methods in computer vision and robotics applications. The major aim of the lecture is to obtain a broad overview of existing methods, and to understand their motivations and main ideas in the context of computer vision and pattern recognition.
|Topic||Lecture Date||Tutorial Date|
|Introduction / Probabilistic Reasoning||20.10||23.10 and 30.10|
|Graphical Models (directed)||3.11||13.11|
|Graphical Models (undirected)||10.11||20.11|
|Bagging and Boosting||24.11||4.12|
|Sequential Data / Hidden Markov Models||1.12||11.12|
|Kernels and Gaussian Processes||8.12||18.12|
|Variational Inference 1||26.1||5.2|
|Variational Inference 2||2.2||5.2|
Linear Algebra, Calculus and Probability Theory are essential building blocks to this course. The homework exercises do not have to be handed in. Solutions for the programming exercises will be provided in Python .
1. Introduction and Probabilistic Reasoning
2. Regression: MLE and MAP
3. Regression \ Directed Graphical Models
4. Undirected Graphical Models
5. Metric Learning
6. Boosting and Bagging
7. HMMs for Sequential Data
8. Kernel Methods and Gaussian Processes
9. Deep Learning
10. GP continued and Clustering I (EM)
11. Clustering II
12. Variational Inference I
13. Variational Inference II (EP)/ Sampling I
14. Sampling II (MCMC)
11. Sampling Methods