Navigation auf uzh.ch

Suche

Department of Informatics Artificial Intelligence and Machine Learning Group

Seminar Artificial Intelligence and Machine Learning

General Information

This seminar is a presentation-only seminar and, thus, no written documents need to be submitted. BA students should prepare talks of 20-30 minutes, MA students of 30 minutes, and a Q&A session of 10-15 minutes will be subsequent to each talk. The Q&A session is open to all participants of the seminar and questions are highly welcome. Presentations will be generally held in English, only BA students may opt to present in German (which is discouraged since non-German speaking students will not be able to follow). Additionally, students will act as hosts for other students' presentations.

Research papers and additional material are handed out three weeks before the presentation so that everyone has the same time to prepare the presentation. Students should prepare their presentation slides approx. one week before their respective seminar and arrange an appointment with their supervisors to discuss them.

An introductory session will be held on the first day of the seminar, where I will introduce and distribute the detailed topics to the students and arrange a schedule for the presentations. Afterward, two presentations will be held in each seminar, the first set of presentations will start two or three weeks after the introductory session, depending on the number of participants. 

The seminar will be held on-site. Presenting students and hosts are supposed to arrive in time to set up the presentation. Participating students are required to be on-site, too, only two of the typically ten or eleven days of presentations can be absent without valid excuse, the usual exemptions apply. A participants list will be need to be signed.

Only in exceptional cases, the seminar will be streamed live via Zoom. There will be no recordings of the presentations.

HS 2024: Deep Learning

The seminar in HS 2024 will focus on the general topic of Deep Learning, including the following sub-topics:

  • Network Architectures
  • Processing of Input Types
  • Areas of Application
  • Analyzing Deep Networks

More details and specific topics will be provided in the fist session of the seminar.

HS 2023: Learning in Humans and Machines

The seminar in HS 2023 will focus on how learning happens in humans and how this is implemented in machine learning techniques. We will focus on four different general topics:

  • Psychology of Learning
  • Learning in Biological Systems
  • Modeling of Human Learning
  • Machine Learning Strategies

More detailed topics will be provided on the first day of the seminar.

HS 2022: Artificial Intelligence for Visual Data Processing

The seminar in HS 2022 will focus on the topics of processing visual data such as images and image-like structures as well as videos via artificial intelligence techniques. Different techniques from simple image filtering to more advanced classification and segmentation will be included, ranging from traditional to modern deep learning approaches, in the following domains:

  • Low-Level Image Processing
  • Complex Image Modeling
  • Image Processing with Deep Learning

More detailed topics will be provided on the first day of the seminar.

HS 2021: Applications of Machine Learning

The seminar in HS 2021 will focus on different applications of machine learning. It will showcase the usage of different machine learning techniques, traditional as well as deep learning techniques, which are applied to solve problems in the following domains:

  • Text Processing
  • Speech Processing
  • Image Processing
  • Robotics
  • Marketing
  • Games

More detailed topics will be provided on the first day of the seminar.

HS 2020: Artificial Neural Networks

The course in HS2020 will focus on artificial neural networks, from the first appearances to the latest trends and observations. Particularly, the seminar will cover the following topics:

  • Network Architectures
  • Loss Functions
  • Input Encodings
  • Spoofing Deep Networks
  • Deep Learning Frameworks

More detailed topics will be provided on the first day of the seminar.