Wahlbereiche / Electives

Collaborative learning, and in particular Federated Learning (FL) is a Machine Learning approach in which multiple clients collaboratively train a Neural Network (NN) model on their private data without the need to share the data. With the increasing large-scale application of FL systems in real-world settings, e.g., for IoT Malware detection or mobile risk management, a number of security, privacy, and functional challenges are posed in the design and implementation of the underlying algorithms and systems.
Besides giving an introduction to the security perspective of Machine Learning, this interactive course will focus on security and privacy attacks and defenses in FL systems. In particular, the course will cover attach such as model/data poisoning and backdoor attacks that allow the adversary to control the outcome of the learning system. The course will cover the introduction of the different components of federated learning systems both on clients and server (aggregation algorithms) and provides hands-on tasks of different difficulty including the state-of-the-art backdoor attacks and defenses.

Advanced type systems offer a compelling way for proving absence of entire classes of errors without even running any program. We will cover among other things the following topics: 

• implementing type systems
• programming with dependent types
• the lambda calculus, simple types, polymorphism, and more
• typing rules, type checking, type inference, type soundness, and more
• syntax, semantics, and mechanization of type systems properties

The course will feature homework using the dependently-typed Lean4, a modern functional and dependently typed programming language, that can be used both to write efficient programs, but at the same time serves as an interactive proof assistant, exploiting the proofs-as-programs correspondence (also known as Curry-Howard).

This is a new course. Before scaling up to at-will registration, participation is limited this semester. Applications will be made on the first lecture day, and selected participants will be notified by/on the second lecture day.

Dieses Seminar ist eine kleine Veranstaltung mit maximal 6 Teilnehmenden. Sie nehmen an zweiwöchentlichen online-Seminarsitzungen teil, erstellen eine multimediale Selbstlernheit und fertigen zusätzlich eine schriftliche Ausarbeitung von ca. 7-10 Seiten an. Das Seminar gehört zu den Modulen "Seminar Angewandte/Praktische Aspekte der Informatik im Unterricht" und "IT in der Grundlehre".

Bei Interesse: Mail an svana.esche@tu-darmstadt.de bis spätestens 16.04.

In this seminar, different security aspects of hardware, systems, artificial intelligence, and embedded devices will be analyzed and discussed. Students will work in groups to process, summarize, and evaluate a number of current scientific publications for a certain topic in the form of an essay. Additionally, each group will present their work at the end of the semester.

Possible topics include:

- Hardware Security
- CPU/Processor Security
- Fuzzing and Automatic Vulnerability Detection
- Security models of current smartphone operating systems (e.g. Android, iOS, Windows Phone, MeeGo, Symbian, RIM)
- Security of mobile devices
- Trusted Hardware
- Internet of Things (IoT)
- Application security (e.g., mobile malware and runtime attacks)
- Privacy aspects in mobile devices
- Security of mobile networks
- Applications of Machine Learning for Security
- Privacy aspects of Deep Neural Networks
- Security Attacks and Defenses against Deep Neural Networks
- Distributed Training of Deep Neural Networks