SpiderBoost and Momentum: Faster Stochastic Variance Reduction Algorithms
This work addresses efficiency and applicability issues in stochastic optimization for machine learning practitioners, offering incremental improvements over existing methods.
The paper tackles the limitations of SPIDER in stochastic variance-reduced algorithms by proposing SpiderBoost, which uses a larger constant stepsize while maintaining near-optimal oracle complexity and handling nonsmooth regularizers via proximal mapping, achieving an improved oracle complexity of O(min{n^{1/2}ε^{-2},ε^{-3}}) in composite nonconvex optimization, with a momentum scheme further accelerating it in experiments.
SARAH and SPIDER are two recently developed stochastic variance-reduced algorithms, and SPIDER has been shown to achieve a near-optimal first-order oracle complexity in smooth nonconvex optimization. However, SPIDER uses an accuracy-dependent stepsize that slows down the convergence in practice, and cannot handle objective functions that involve nonsmooth regularizers. In this paper, we propose SpiderBoost as an improved scheme, which allows to use a much larger constant-level stepsize while maintaining the same near-optimal oracle complexity, and can be extended with proximal mapping to handle composite optimization (which is nonsmooth and nonconvex) with provable convergence guarantee. In particular, we show that proximal SpiderBoost achieves an oracle complexity of $\mathcal{O}(\min\{n^{1/2}ε^{-2},ε^{-3}\})$ in composite nonconvex optimization, improving the state-of-the-art result by a factor of $\mathcal{O}(\min\{n^{1/6},ε^{-1/3}\})$. We further develop a novel momentum scheme to accelerate SpiderBoost for composite optimization, which achieves the near-optimal oracle complexity in theory and substantial improvement in experiments.