ABSTRACT
We investigate the zero- and finite-temperature properties of the random-bond s=1/2 Heisenberg antiferromagnet on the pyrochlore lattice by the exact diagonalization and the Hams-de Raedt methods. We find that the randomness induces the gapless quantum spin liquid (QSL) state, the random-singlet state. Implications to recent experiments on the mixed-anion pyrochlore-lattice antiferromagnet Lu_{2}Mo_{2}O_{5}N_{2} exhibiting gapless QSL behaviors are discussed.
ABSTRACT
The nature of the randomness-induced quantum spin liquid state, the random-singlet state, is investigated in two dimensions (2D) by means of the exact-diagonalization and the Hams-de Raedt methods for several frustrated lattices, e.g. the triangular, the kagome and the J 1-J 2 square lattices. Properties of the ground state, the low-energy excitations and the finite-temperature thermodynamic quantities are investigated. The ground state and the low-lying excited states consist of nearly isolated singlet-dimers, clusters of resonating singlet-dimers, and orphan spins. Low-energy excitations are either singlet-to-triplet excitations, diffusion of orphan spins accompanied by the recombination of nearby singlet-dimers, creation or destruction of resonating singlet-dimers clusters. The latter two excitations give enhanced dynamical 'liquid-like' features to the 2D random-singlet state. Comparison is made with the random-singlet state in a 1D chain without frustration, the similarity and the difference between in 1D and in 2D being highlighted. Frustration in a wide sense, not only the geometrical one but also including the one arising from the competition between distinct types of interactions, play an essential role in stabilizing this frustrated random singlet state. Recent experimental situations on both organic and inorganic materials are reviewed and discussed.
ABSTRACT
Statistical optimality in multipartite ranking is investigated as an extension of bipartite ranking. We consider the optimality of ranking algorithms through minimization of the theoretical risk which combines pairwise ranking errors of ordinal categories with differential ranking costs. The extension shows that for a certain class of convex loss functions including exponential loss, the optimal ranking function can be represented as a ratio of weighted conditional probability of upper categories to lower categories, where the weights are given by the misranking costs. This result also bridges traditional ranking methods such as proportional odds model in statistics with various ranking algorithms in machine learning. Further, the analysis of multipartite ranking with different costs provides a new perspective on non-smooth list-wise ranking measures such as the discounted cumulative gain and preference learning. We illustrate our findings with simulation study and real data analysis.
