University of Crete

About: University of Crete is a education organization based out in Rethymno, Greece. It is known for research contribution in the topics: Population & Galaxy. The organization has 8681 authors who have published 21684 publications receiving 709078 citations. The organization is also known as: Panepistimio Kritis.

Quantum stability of a w − 1 phase of cosmic acceleration

Abstract: We consider a massless, minimally coupled scalar with a quartic self-interaction which is released in Bunch-Davies vacuum in the locally de Sitter background of an inflating universe. It was shown, in this system, that quantum effects can induce a temporary phase of superacceleration, causing a violation of the weak energy condition on cosmological scales. In this paper, we investigate the system's stability by studying the behavior of linearized perturbations in the quantum-corrected effective field equation at one- and two-loop order. We show that the amplitude of the quantum-corrected mode function is reduced in time, starting from its initial classical (Bunch-Davies) value. This implies that the linear perturbations do not grow; hence, the model is stable. The decrease in the amplitude is in agreement with the system developing a positive (growing) mass squared due to quantum processes. The induced mass, however, remains perturbatively small and does not go tachyonic. This ensures the stability.

From K-Means to Higher-Way Co-Clustering: Multilinear Decomposition With Sparse Latent Factors

Abstract: Co-clustering is a generalization of unsupervised clustering that has recently drawn renewed attention, driven by emerging data mining applications in diverse areas. Whereas clustering groups entire columns of a data matrix, co-clustering groups columns over select rows only, i.e., it simultaneously groups rows and columns. The concept generalizes to data “boxes” and higher-way tensors, for simultaneous grouping along multiple modes. Various co-clustering formulations have been proposed, but no workhorse analogous to K-means has emerged. This paper starts from K-means and shows how co-clustering can be formulated as a constrained multilinear decomposition with sparse latent factors. For three- and higher-way data, uniqueness of the multilinear decomposition implies that, unlike matrix co-clustering, it is possible to unravel a large number of possibly overlapping co-clusters. A basic multi-way co-clustering algorithm is proposed that exploits multilinearity using Lasso-type coordinate updates. Various line search schemes are then introduced to speed up convergence, and suitable modifications are proposed to deal with missing values. The imposition of latent sparsity pays a collateral dividend: it turns out that sequentially extracting one co-cluster at a time is almost optimal, hence the approach scales well for large datasets. The resulting algorithms are benchmarked against the state-of-art in pertinent simulations, and applied to measured data, including the ENRON e-mail corpus.