Conference blog.

I am attending Thirty-seventh Conference on Neural Information Processing Systems (Neurips2023) now and I am happy to share my experience with you all. I will write a short summary every day.

Workshop

It was a difficult choice for me to pick one particular workshop to attend, like yesterday. But since I am researching on machine learning theory, the workshop on Mathematics of Modern Machine Learning seems the first choice for me.

The talk, which is the highlight of my day, is Benefits of learning with symmetries: eigenvectors, graph representations and sample complexity. TLDR: Many datasets or tasks have some group-equivariant structure. Exploiting this one could get a refined generalization bound. The talk is delivered elegantly with mathematics and examples.

Postface

The conference is ended and it has given me a revelation on machine learning. I am really glad that I have attended the conference. Now it is time to drink a cup of tea and digest what I have learnt in New Orleans.