On the Benefits of Convolutional Models: a Kernel Perspective
Speaker: Alberto Bietti, NYU Center for Data Science
Location: 60 Fifth Avenue 150
Date: April 19, 2022, 2 p.m.
Host: Joan Bruna
Many supervised learning problems involve high-dimensional data such as images, text, or graphs. In order to make efficient use of data, it is often useful to leverage priors in the problem at hand, such as invariance to certain transformations or stability to small deformations. Empirically, deep convolutional networks have been very successful on such data, raising the question of how they are able to capture relevant structure in these problems for efficient learning.
My work studies this question from a theoretical perspective using kernel methods, in particular convolutional kernels. These are constructed following similar architectural principles as convolutional networks, are closely related to their infinite-width limits in certain regimes, and provide good empirical performance on standard computer vision benchmarks such as Cifar10. I will present contributions that highlight the benefits of (deep) convolutional architectures in terms of stability to deformations and sample complexity.
Alberto Bietti is currently a faculty fellow at the NYU Center for Data Science. He received his PhD in 2019 from Inria and Université Grenoble-Alpes, working under the supervision of Julien Mairal, and was awarded the best PhD prize from Université Grenoble-Alpes for his thesis work. In 2020, he was a postdoc at Inria Paris hosted by Francis Bach. His research spans machine learning, optimization, and statistics, with a focus on building the theoretical foundations of deep learning. Before his PhD, he worked as a software engineer at Quora.
In-person attendance only available to those with active NYU ID cards.