Computational Geometry & Topology Challenge 2021

Modern real-world data typically have high dimensional ambient measurement spaces. Yet, they often exhibit low dimensional intrinsic structures that can be uncovered by geometry- or topology- oriented methods. As a result, recent years have seen significant interest and progress in geometric and topological approaches to representation learning.

The purpose of this challenge is to push forward the fields of computational differential geometry and topology, by creating the best data analysis, computational method, or numerical experiment relying on state-of-the-art geometric and topological Python packages. All participants will have the opportunity to co-author a white paper summarizing the findings of the challenge.

For more details, visit the GitHub repository of the challenge: https://github.com/geomstats/challenge-iclr-2021.

The challenge is over now! Please read The Official ICLR 2021 Challenge for Computational Geometry & Topology: Design and Results white paper for details on the submissions.

Deadline:

May 2nd, 2021 at 23:59 PST (Pacific Standard Time).

Announcement of winners and prizes:

The first 3 winners will be announced at the ICLR 2021 Virtual Workshop on Geometrical & Topological Representation Learning and advertised through the web.

The prizes are:

• 2000$ for the 1st place,
• 1000$ for the 2nd place,
• 500$ for the 3rd place.

We are looking forward to your contributions!

Results:

All submissions to the challenge were of very high quality. We would like to first congratulate all of the participants! We encourage the readers to take a look at the participants’ notebooks on GitHub: https://github.com/geomstats/challenge-iclr-2021.

The winners were decided using the Condorcet Method.

We are now excited to announce the best contributions to open-source computational geometry and topology of ICLR 2021!

Winner 1st place ($2000):

Matei Bejan, Bogdan Luchian, Andrei Manolache, Armand Nicolicioiu, Mihaela Nistor, Mihai Sorin Stupariu

Topological noise invariant features using Giotto-TDA and Geomstats

Co-winners for 2nd and 3rd place ($750 each):

• Saiteja Utpala: Shrinkage Estimator of SPD matrices & Recursive Karcher Expectation Estimator
• Elodie Maignant: Visualization of Kendall shape spaces for triangles

Special mentions:

• Gabriele Corso: Neural Sequence Distance Embeddings
• Somesh Mohapatra: Repurposing peptide inhibitors for SARS-Cov-2 spike protein

Congratulations to all!!

The organizers will be in touch with the winners to distribute the prizes, and in touch with all the participants to write the white paper summarizing the findings of the challenge.