4212 Deep Learning for IT-Security
FT, MSc Cyber-Security
6 ECTS, Lecture + Seminar
Link to the Course in ILIAS - TBA
In the last decade, the availability of larger amounts of data and higher computational resources has led to a revolution in machine learning, leading to the so called deep learning architectures. These algorithms can provide higher classification accuracy or generate very realistic synthetic images or audios.
This course begins with a review of the main shortcomings of traditional machine learning algorithms as a motivation for deep learning (DL) architectures. The fundamental building blocks of DL algorithms are be presented: the evolution from the simple perceptron to multi-layer neural networks, as well as techniques covering regularisation, normalisation, and optimisation for training these large networks. Specialised architectures for computer vision (e.g., CNNs) and sequence processing (e.g., LSTMs), generative AI (e.g., GANs), and multi-input/output networks are the focus of the subsequent chapters. Practical examples in Python are introduced during the course in order to provide the students with hands-on experience. Furthermore, examples on the use of these architectures for IT-security related applications are discussed.
In the seminar, students will be able to implement and evaluate current deep learning architectures. Students will analyse, evaluate and discuss current challenges of deep learning and its applications in IT security in order to find new solutions. To do this, they will research specialised literature and current publications.