Skip to content. | Skip to navigation

Personal tools

Theme for TIFR Centre For Applicable Mathematics, Bangalore


You are here: Home / Events / SGD from the Lens of Markov process: An Algorithmic Stability Perspective

SGD from the Lens of Markov process: An Algorithmic Stability Perspective

Speaker: Anant Raj (SIERRA Project team, Inria/CSL, UIUC)
Speaker: Anant Raj (SIERRA Project team, Inria/CSL, UIUC)
When Nov 28, 2023
from 04:00 PM to 05:00 PM
Where Lecture Hall- LH 111 (Hybrid)
Add event to calendar vCal

Abstract: In this talk, we delve into the intricate relationships between heavy-tailed distributions, generalization error, and algorithmic stability in the realm of noisy stochastic gradient descent. Recent research has illustrated the emergence of heavy tails in stochastic optimization and their intriguing links to generalization error. However, these studies often relied on challenging topological and statistical assumptions. Empirical evidence has further challenged existing theory, suggesting that the relationship between heavy tails and generalization is not always monotonic. In response, we introduce novel insights, exploring the relationship between tail behavior and generalization properties through the lens of algorithmic stability. Our analysis reveals that the stability of stochastic gradient descent (SGD) varies based on how we measure it, leading to interesting conclusions about its behavior. Expanding upon these findings, we extend the scope to a broader class of objective functions, including non-convex ones. Leveraging Wasserstein stability bounds for heavy-tailed stochastic processes, our research sheds light on the non-monotonic connection between generalization error and heavy tails, offering a more comprehensive perspective. Additionally, we introduce a unified approach for proving Wasserstein stability bounds in stochastic optimization, emphasizing time-uniform stability and its role in various scenarios, including convex and non-convex losses. Our approach is versatile and applicable to popular optimizers, highlighting the importance of ergodicity.

Join Zoom Meeting

Meeting ID: 965 8981 5643
Passcode: 877731

Filed under: