REVIEW 2 cited by
High Probability Complexity Bounds for Non-Smooth Stochastic Optimization with Heavy-Tailed Noise
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Stochastic first-order methods are standard for training large-scale machine learning models. Random behavior may cause a particular run of an algorithm to result in a highly suboptimal objective value, whereas theoretical guarantees are usually proved for the expectation of the objective value. Thus, it is essential to theoretically guarantee that algorithms provide small objective residual with high probability. Existing methods for non-smooth stochastic convex optimization have complexity bounds with the dependence on the confidence level that is either negative-power or logarithmic but under an additional assumption of sub-Gaussian (light-tailed) noise distribution that may not hold in practice. In our paper, we resolve this issue and derive the first high-probability convergence results with logarithmic dependence on the confidence level for non-smooth convex stochastic optimization problems with non-sub-Gaussian (heavy-tailed) noise. To derive our results, we propose novel stepsize rules for two stochastic methods with gradient clipping. Moreover, our analysis works for generalized smooth objectives with H\"older-continuous gradients, and for both methods, we provide an extension for strongly convex problems. Finally, our results imply that the first (accelerated) method we consider also has optimal iteration and oracle complexity in all the regimes, and the second one is optimal in the non-smooth setting.
Forward citations
Cited by 2 Pith papers
-
High-Probability Last-Iterate Guarantees for Two-Point Gaussian Zeroth-Order Stochastic Gradient Descent
Same-sample two-point Gaussian ZO-SGD achieves Õ(d/T) last-iterate suboptimality with probability 1−δ under conditional sub-Gaussian noise, with only logarithmic 1/δ dependence.
-
Tight Long-Term Tail Decay of (Clipped) SGD in Non-Convex Optimization
For non-convex smooth costs, the tail probability that SGD's best gradient remains above a fixed threshold decays at speed t/log(t) (bounded noise), and clipped SGD achieves t^{4(p-1)/(3p-2)}/log(t) under p-th moment ...
Discussion (0). Continue with ORCID to comment.