After successful completion of the course, students are able to understand basic methods in the field of machine learning and to solve typical problems by applying sophisticated machine learning algorithms.
The course provides a practical and theoretical overview of machine learning methods: General concepts of machine learning (supervised vs. unsupervised approaches, generalization and overfitting, regularization techniques, evaluation methods, cross-validation, early stopping). Basic regression and classification algorithms (least squares, perceptron and logistic regression). Basics of clustering (Lloyd, kMeans). Nonlinear extensions through the kernel trick (support vector machines, kernel ridge). Gradient descent and neural networks (backprogation, LMS, RLS, Newton). Domain specific layers (Recursive and convolutional networks) and unsupervised neural networks (autoencoders).
The lectures are complemented by 4 exercise classes with a focus on practical implementations in Tensorflow and Numpy. Basic Python knowledge is beneficial but not required. At the end of the course the students will work on a practical project, undergoing a complete machine learning pipeline on a real-world dataset. Here we also cover more advanced topics (adversarial networks, transfer learning).
The first lecture will be held in presence in EI1 on October 6, 2023, 2pm-3:30pm.
This years exercise classes will be held on:
20.10.23
24.11.23
15.12.23
12.01.24
Students gain points in the 4 exercises classes by submitting homeworks and presenting their solutions. The obligatory project report handed in at the end of the semester acts as the basis for the final oral exam.
Not necessary
SP1 and 2 should be an active knowledge.
