Abstract— The first step toward development a “nonsmooth” version of Speed-Gradient (SG) method is made. As a basis for.

Nonsmooth versions of SG-algorithm in differential and finite form are formulated. Conditions for the control goal achievement are proposed. The results are ...

In this article, nonsmooth extensions of the Speed-Gradient (SG) algorithms in differential and finite forms are proposed.

Jan 16, 2023 · This paper proposes a more efficient algorithm using stochastic recursive gradient estimators, which improves the complexity to \mathcal{O}(L^3 ...

Nonsmooth versions of SG-algorithm in differential and finite form are formulated. Conditions for the control goal achievement are proposed. The results are ...

The first step toward development a “nonsmooth” version of Speed-Gradient (SG) method in differential and finite form is made and the Hadamard ...

In this paper, we propose an efficient stochastic gradient-free method named GFM+ for nonsmooth nonconvex stochastic optimization.

The proposed finite-differential algorithms possess enhanced dynamic properties and provide extra flexibility for control system design. Particularly, they may ...

Nonsmooth versions of SG-algorithm in differential and finite form are formulated. Conditions for the control goal achievement are proposed. The results are ...

We propose and analyze several stochastic gradient algorithms for finding stationary points or local minimum in nonconvex, possibly with nonsmooth ...