arXiv subjects

Cheng Tang

Publications and source records attributed to Cheng Tang.

57 records · Page 4Linked to original sources

arXiv2017

Eigenvalue and Eigenfunction for the $PT$-symmetric Potential $V = - (ix)^N$

The real energy spectrum from the $PT$-symmetric Hamiltonian $H = p^2 - (ix)^N$ with $x\in\mathbb{C}$ was examined within one pair of Stokes wedges in 1998 by Bender and Boettcher. For this Hamiltonian we discuss the following three questions. First, since their paper used a Runge-Kutta method to integrate along a path at the center of the Stokes wedges to calculate eigenvalues $E$ with high accuracy, we wonder if the same eigenvalues can be obtained if integrate along some other paths in different shapes. Second, what the corresponding eigenfunctions look like? Should the eigenfunctions be independent from the shapes of path or not? Third, since for large $N$ the Hamiltonian contains many pairs of Stokes wedges symmetric with respect to the imaginary axis of $x$, thus multiple families of real energy spectrum can be obtained. What do they look like? Any relation among them?

math-ph↗

arXiv2016

Convergence rate of stochastic k-means

We analyze online \cite{BottouBengio} and mini-batch \cite{Sculley} $k$-means variants. Both scale up the widely used $k$-means algorithm via stochastic approximation, and have become popular for large-scale clustering and unsupervised feature learning. We show, for the first time, that starting with any initial solution, they converge to a "local optimum" at rate $O(\frac{1}{t})$ (in terms of the $k$-means objective) under general conditions. In addition, we show if the dataset is clusterable, when initialized with a simple and scalable seeding algorithm, mini-batch $k$-means converges to an optimal $k$-means solution at rate $O(\frac{1}{t})$ with high probability. The $k$-means objective is non-convex and non-differentiable: we exploit ideas from recent work on stochastic gradient descent for non-convex problems \cite{ge:sgd_tensor, balsubramani13} by providing a novel characterization of the trajectory of $k$-means algorithm on its solution space, and circumvent the non-differentiability problem via geometric insights about $k$-means update.

cs.LG↗

arXiv2016

Convergence rate of stochastic k-means

We analyze online and mini-batch k-means variants. Both scale up the widely used Lloyd 's algorithm via stochastic approximation, and have become popular for large-scale clustering and unsupervised feature learning. We show, for the first time, that they have global convergence towards local optima at $O(\frac{1}{t})$ rate under general conditions. In addition, we show if the dataset is clusterable, with suitable initialization, mini-batch k-means converges to an optimal k-means solution with $O(\frac{1}{t})$ convergence rate with high probability. The k-means objective is non-convex and non-differentiable: we exploit ideas from non-convex gradient-based optimization by providing a novel characterization of the trajectory of k-means algorithm on its solution space, and circumvent its non-differentiability via geometric insights about k-means update.

cs.LG↗