Abstract
The consistency of a non-local anisotropic Ginzburg–Landau type functional for data classification and clustering is studied. The Ginzburg–Landau objective functional combines a double well potential, that favours indicator valued functions, and the p-Laplacian, that enforces regularity. Under appropriate scaling between the two terms, minimisers exhibit a phase transition on the order of ɛ = ɛn, where n is the number of data points. We study the large data asymptotics, i.e. as n → ∝, in the regime where ɛn → 0. The mathematical tool used to address this question is Γ-convergence. It is proved that the discrete model converges to a weighted anisotropic perimeter.
Publisher
Cambridge University Press (CUP)
Reference57 articles.
1. A non-local anisotropic model for phase transitions: asymptotic behaviour of rescaled energies
2. Consistency of Cheeger and ratio graph cuts;García Trillos;J. Mach. Learn. Res.,2016
3. The effect of a singular perturbation on nonconvex variational problems
4. [50] Szlam, A. & Bresson, X. (2009) A total variation-based graph clustering algorithm for cheeger ratio cuts. Technical report UCLA CAM Report 9–68, Los Angeles, pp. 1–12.
5. [53] Thorpe, M. & Slepčev, D. (2017, in preparation) Transportation Lp distances: Properties and extensions.
Cited by
5 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献