OCLGMLJan 21

TRSVR: An Adaptive Stochastic Trust-Region Method with Variance Reduction

arXiv:2601.14647v1h-index: 4
Originality Incremental advance
AI Analysis

This work addresses optimization efficiency for machine learning practitioners, but it is incremental as it builds on existing SVRG and trust-region techniques.

The authors tackled the problem of accelerating convergence in unconstrained nonconvex optimization by proposing a stochastic trust-region method with variance reduction, which achieves iteration and sample complexity bounds matching SVRG-based methods and outperforms SGD and Adam on machine learning tasks.

We propose a stochastic trust-region method for unconstrained nonconvex optimization that incorporates stochastic variance-reduced gradients (SVRG) to accelerate convergence. Unlike classical trust-region methods, the proposed algorithm relies solely on stochastic gradient information and does not require function value evaluations. The trust-region radius is adaptively adjusted based on a radius-control parameter and the stochastic gradient estimate. Under mild assumptions, we establish that the algorithm converges in expectation to a first-order stationary point. Moreover, the method achieves iteration and sample complexity bounds that match those of SVRG-based first-order methods, while allowing stochastic and potentially gradient-dependent second-order information. Extensive numerical experiments demonstrate that incorporating SVRG accelerates convergence, and that the use of trust-region methods and Hessian information further improves performance. We also highlight the impact of batch size and inner-loop length on efficiency, and show that the proposed method outperforms SGD and Adam on several machine learning tasks.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes