piwik-script

Deutsch Intern
    Astronomy

    Data Lab

    DataSphere@JMUW

    Co-working space with a GPU/CPU compute-cluster for researchers working on machine-learning problems

    The Chair of Astronomy  hosts the data lab Datasphere@JMUW as a meeting room for researchers working on machine-learning problems in data-driven science.  The data lab is located on Campus Hubland North, Emil-Fischer-Str. 31, 1st floor, in room 31.01.007. The data lab is equipped with a powerful GPU-cluster to support its machine-learning user community.

    Access to the lab will be possible starting with the summer term of 2020.  Please, describe your project and provide the name of your scientific advisor when you request access.  The opportunity targets participants of ML-group at the Chair of Astronomy and the interdisciplinary seminar on maschine-learning use-cases.

    Requests can be addressed to our team coach Paul Ray Burd (contact: pburd"@"astro.uni-wuerzburg.de, phone 8711) who also registers the participants of the seminar.

    The data lab Datasphere@JMUW is a joint project of the Chairs of Astronomy (Prof. Mannheim) und Data Science (Prof. Hotho) at the crossroads between data-driven sciences and the AI-centre CAIDAS of the University. The project is funded by DATEV Foundation Future.

    COVID-19:

    • The ML-seminar will be continued as a Zoom-Videomeeting.  Registered participants will receive invitations in due time.  Suggestions for talks are always welcome and should be directed to Paul Burd by Email.
    • A virtual mirror of the data lab has been set up as a Slack-Workspace (astro-wue).  Details on how to use the GPU-cluster remotely will be deployed here.

     

    Thursdays 16:15 Uhr, Raum 31.00.017

     

     
    Date Speaker

    Title

    07.03.2019 Paul Ray Burd Radio frequency interference filtering with LSTM cells
    25.07.2019 Martin Blaimer

    Magnetic-Resonance-Imaging (MRI): Sequence optimization with Machine Learning

    20.02.2020

    Peter Dawood

    Neural Networks for Magnetic Resonance Imaging Reconstructions

     

    Rick Seifert

    DeepCLEM: automated Registration of correlative light- and electron microscopy Images with convolutional neural networks

    27.02.2020

    Jannik Stebani

    Enhancing quantitative Magnetic Resonance Fingerprinting via deep neural nets

      Andreas Berberich Generative Adversarial Networks for single-molecule localization microscopy