|
|
3
|
|
44223
|
Lecture-Discussion
|
A1
|
10:00AM
-10:50AM
|
MWF
|
106B6 Engineering Hall
|
Wang, S
|
- Availability:
- Open (Restricted)
- Part of Term:
- 1
- Date Range:
- 08/24/26-12/09/26
- Section Title:
- Topic
- Section Info:
- Statistical Foundations of Self-Supervised Learning Description: This course provides a rigorous introduction to Self-Supervised Representation Learning (SSL), focusing on bridging the gap between empirical deep learning and formal statistical theory. We will explore how models learn representations from unlabeled data by solving pretext tasks. The curriculum is structured around two central themes: first, a comprehensive survey of the algorithmic landscape, examining state-of-the-art frameworks such as Contrastive Methods, Masked Modeling, Video Representation Learning, and Multi-Modal Alignment; and second, an exploration of statistical foundations and recent progress. In this latter half, we will discuss the evolving theoretical underpinnings of these methods, with a focus on understanding the role of self-supervised signals, the geometry of latent spaces, and the statistical properties that allow these representations to transfer effectively to downstream tasks. The section prerequisites are STAT 426 and STAT 511/510. For Statistics course registration information: go.illinois.edu/StatisticsRegistration.
- Restriction(s):
-
Restricted to Graduate - Urbana-Champaign.
Restricted to students in the Statistics department.
|