The Department of Computer Science at the University of Copenhagen offers a HYBRID course “Machine Learning A”. It is a 7.5 ECTS master-level course introducing basic theory and algorithms of machine learning. The course will run in Sep-Oct 2024 in a hybrid format and will support fully remote participation. The lectures will be streamed via Zoom and recorded, and the students will have a choice of physical and online exercise classes. The course welcomes students from other universities and people from industry.
The course assumes solid knowledge of linear algebra, calculus, and probability theory, and basic Python programming skills. It also assumes basic knowledge of LaTeX for typing home assignments.
Topics to be covered:
The meaning of generalization beyond a finite sample under the independent identically distributed (i.i.d.) assumption
K-Nearest Neighbors algorithm
Perceptron
Linear Regression
Logistic Regression
Regularization and working in the feature space
Markov's, Chebyshev's, and Hoeffding's inequalities
Generalization bounds based on training/validation/test sets
Lower bound for generalization (the impossibility of generalization in the worst case)
Occam's Razor bound for countable sets of prediction rules and its application to Decision Trees
Random Forests
Neural Networks
Principal Component Analysis (PCA)
Clustering: k-means & k-means++ algorithms
Non-linear Dimensionality Reduction via Stochastic Neighbor Embedding
Further information about the course: https://kurser.ku.dk/course/ndak22000u/ and https://sites.google.com/diku.edu/machine-learning-courses/mla.
Registration links:
Contact person: Sadegh Talebi <mstale...@gmail.com>
Registration deadline - 27 August