Data Science Chair

    XtAI Lab

    XtAI Lab I, II, III

    Lecturer: Prof. Dr. Andreas Hotho
    Contact: Janna Omeliyanenko

    Credits: 5 ECTS or 10 ECTS (credits can be adapted by arrangement)

       Introductory event XtAI Labs:  18.10.2023, 10:15 am,  HS4 (Lecture Hall 4) the Science Lecture Hall Building (P4) 

    Further chairs offering XtAI Labs:

    Introductory Event

    At the XtAI Labs introductory event, chairs IV, IX, X , XV and Reinforcement Learning and Computational Decision-Making chair will provide information on XtAI Labs and topics offered this winter term.

    Date: 18.10.2023, 10:15 am
    Location: HS4 (Lecture Hall 4) the Science Lecture Hall Building (P4)

    XtAI Labs concept

    The XtAI Labs provides knowledge about the most important steps and tools for the design and development of an XtAI application. Knowledge such as common data handling and processing techniques, libraries, and integration with augmented reality applications will be taught in a theoretical or hands-on format. Group work will include planning, design, creation, evaluation, and refinement of a comprehensive XtAI application models.

    Current tasks

    • Group Project 1: Knowledge Graph Embedding with Language Models 
      • Assess the quality of state-of-the-art graph embedding models with "Link Prediction".
    • Group Project 2: Parameter Efficient Smart Language Models
      • Building an intelligent language model while controlling the appropriate parameter size.

    For detailed information on the tasks, please contact Ms. Janna Omeliyanenko


    To register for the XtAI Lab please email to Ms. Janna Omeliyanenko  with the following information:

    • name of task you would like to work on
    • number of participants in your group 

    Registration deadline: will be announced at the introductory event on 18.10.2023, 10:15 am

    Performance review

    At the end of the semester, each group will present their approach in a 30 minute presentation. In addition, an lab report of 20 pages must be submitted.