Pith. sign in

Stochastic Three Points Method for Unconstrained Smooth Minimization

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it
abstract

In this paper we consider the unconstrained minimization problem of a smooth function in ${\mathbb{R}}^n$ in a setting where only function evaluations are possible. We design a novel randomized derivative-free algorithm --- the stochastic three points (STP) method --- and analyze its iteration complexity. At each iteration, STP generates a random search direction according to a certain fixed probability law. Our assumptions on this law are very mild: roughly speaking, all laws which do not concentrate all measure on any halfspace passing through the origin will work. For instance, we allow for the uniform distribution on the sphere and also distributions that concentrate all measure on a positive spanning set. Given a current iterate $x$, STP compares the objective function at three points: $x$, $x+\alpha s$ and $x-\alpha s$, where $\alpha>0$ is a stepsize parameter and $s$ is the random search direction. The best of these three points is the next iterate. We analyze the method STP under several stepsize selection schemes (fixed, decreasing, estimated through finite differences, etc). We study non-convex, convex and strongly convex cases.

fields

cs.LG 1

years

2026 1

verdicts

UNVERDICTED 1

representative citing papers

Finding Stationary Points by Comparisons

cs.LG · 2026-06-25 · unverdicted · novelty 6.0

Presents classical Õ(n²/ε^{1.5}) and quantum Õ(n/ε^{1.5}) query algorithms for ε-stationary points of twice-differentiable non-convex functions with Lipschitz gradient and Hessian via comparison oracles.

citing papers explorer

Showing 1 of 1 citing paper.

  • Finding Stationary Points by Comparisons cs.LG · 2026-06-25 · unverdicted · none · ref 5 · internal anchor

    Presents classical Õ(n²/ε^{1.5}) and quantum Õ(n/ε^{1.5}) query algorithms for ε-stationary points of twice-differentiable non-convex functions with Lipschitz gradient and Hessian via comparison oracles.