Year 2021 in Review & Predictions for 2022

What does 2022 hold for Geometric & Graph ML?

Last year, I sought the opinion of leading researchers of Graph ML to make predictions about the future development in the field. This year, we teamed up with Petar Veličković and interviewed a cohort of distinguished and prolific experts in an attempt to summarise the highlights of the past year and predict what is in store for 2022.

Image: Shutterstock

This post was co-authored with Petar Veličković. See also my last year’s prediction, Michael Galkin’s excellent post on the current state of affairs in Graph ML, a deeper dive into subgraph GNNs, techniques inspired by PDEs and differential geometry and algebraic topology, and how the concepts of symmetry and invariance form the picture of modern deep learning.

Summing up impressions of 2021 and forecasting the year ahead with the help of the leading domain experts is a worthwhile and rewarding experiment, as it gives a range of diverse insights and opinions to learn from. In the words of one of our interviewees, who keeps an attentive eye on the top emerging trends across the entire machine learning landscape:

“With geometric and graph-based ML methods going from niche to one of the hottest fields of AI research today, it is exciting to imagine what the next twelve months will hold.” — Nathan Benaich, General Partner of Air Street Capital and co-author of the State of AI Report

In retrospect, this effort turned out not as easy a feat as we have initially thought, since the field has hugely grown in the past year — this is witnessed by the length of this post, the longest we have written so far.

Quick take-home messages (if you are lazy to read the rest)

1. Geometry becomes increasingly important in ML. Differential geometry and cognate fields have brought new…