arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1703.01339math.OCmath.DS
收藏

非凸最优化问题的类牛顿动力学

Newton-like dynamics associated to nonconvex optimization problems

Radu Ioan Bot, Ernö Robert Csetnek

摘要:

考虑动力系统\begin{equation*}\left\{ \begin{array}{ll} v(t)\in\partialϕ(x(t))\\ λ\dot x(t) + \dot v(t) + v(t) + \nabla ψ(x(t))=0, \end{array}\right.\end{equation*},其中$ϕ:\R^n\to\R\cup\{+\infty\}$是一个正常的、凸的、下连续的函数,$ψ:\R^n\to\R$是一个(可能非凸的)光滑函数,$λ>0$是一个控制速度的参数.我们表明,轨迹$x$的极限点集包含在目标函数$ϕ+ψ$的临界点集,这是在这里被视为其极限次微分的零点集。如果目标函数满足Kurdyka-Schojasiewicz性质,则可以证明整个轨迹$x$收敛到一个临界点。此外,轨道的收敛速度得到的目标函数的Chaojasiewicz指数,提供后者满足Chaojasiewicz性质。

英文摘要:

We consider the dynamical system \begin{equation*}\left\{ \begin{array}{ll} v(t)\in\partialϕ(x(t))\\ λ\dot x(t) + \dot v(t) + v(t) + \nabla ψ(x(t))=0, \end{array}\right.\end{equation*} where $ϕ:\R^n\to\R\cup\{+\infty\}$ is a proper, convex and lower semicontinuous function, $ψ:\R^n\to\R$ is a (possibly nonconvex) smooth function and $λ>0$ is a parameter which controls the velocity. We show that the set of limit points of the trajectory $x$ is contained in the set of critical points of the objective function $ϕ+ψ$, which is here seen as the set of the zeros of its limiting subdifferential. If the objective function satisfies the Kurdyka-Łojasiewicz property, then we can prove convergence of the whole trajectory $x$ to a critical point. Furthermore, convergence rates for the orbits are obtained in terms of the Łojasiewicz exponent of the objective function, provided the latter satisfies the Łojasiewicz property.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑