ML Reproducibility challenge 2022 Scalable Machine Learning


This project is based on the Machine Learning Reproducibility Challenge:
In this project, your team will participate in the challenge by reproducing the contribution of a paper from some of the major machine learning conferences last year (NeurIPS/ICML/ICLR etc.).
You can choose the paper from within a list of papers we provide.
You will be able to take a deep dive into the paper and get hands-on experience implementing state-of-the-art techniques in the paper. You will also get a glimpse into the world of scientific research and publication.

Sign up

Expected results
  • Developing an in-depth understanding of the selected paper
  • Writing and running code to reproduce the main claims of the
  • Reporting your work in the form of a reproducibility report, which describes:
    • The target questions of your report,
    • Experimental methodology,
    • Implementation details,
    • Analysis and discussion of findings,
    • Conclusions on reproducibility of the paper
Beyond our expectations
  • Based on your code and in-depth understanding, you might be able to produce additional insights in the paper’s results
  • Well-written and polished reproducibility reports can be submitted in response to the challenge.
  • Submitted reports will be peer-reviewed and then published online alongside the original papers.
  • Exceptional reports might even be published in the ReScience academic journal
Course Participants

Both Bachelor’s and Master’s students are welcome to participate! Bachelor students will get 8CP, Master students 10CP.
The project will be conducted in groups:

  • Students per group: 1 to 3
  • Max. number of groups: 4 to 6
  • Max. total students: 12

You can sign up in pairs or be assigned a partner when the project starts

Required Skills

To sign up for this course, please make sure you have:

  • Basic Machine Learning knowledge
  • Familiarity with Python
  • Familiarity with at least one Deep Learning Framework (PyTorch, Tensorflow, Jax, etc.)

When you sign up for the project, please provide a short summary (one sentence per point) of your skills and prior experience for each of the points above.

Nice to have (but not required):

Attended one or more of the following courses: Introduction to Artificial Intelligence, Maschinelles Lernen, Machine Learning: Supervised Methods, Artificial Neural Networks, Machine Learning: Evolutionary Algorithms, Machine Learning: Unsupervised Methods, Topics in Deep Learning for Sequence Processing

The Institut für Neuroinformatik (INI) is a central research unit of the Ruhr-Universität Bochum. We aim to understand the fundamental principles through which organisms generate behavior and cognition while linked to their environments through sensory systems and while acting in those environments through effector systems. Inspired by our insights into such natural cognitive systems, we seek new solutions to problems of information processing in artificial cognitive systems. We draw from a variety of disciplines that include experimental approaches from psychology and neurophysiology as well as theoretical approaches from physics, mathematics, electrical engineering and applied computer science, in particular machine learning, artificial intelligence, and computer vision.

Universitätsstr. 150, Building NB, Room 3/32
D-44801 Bochum, Germany

Tel: (+49) 234 32-28967
Fax: (+49) 234 32-14210