DeepCF: A Unified Framework of Representation Learning and Matching Function Learning in Recommender System

In general, recommendation can be viewed as a matching problem, i.e., match proper items for proper users. However, due to the huge semantic gap between users and items, it’s almost impossible to directly match users and items in their initial representation spaces. To solve this problem, many methods have been studied, which can be generally categorized into two types, i.e., representation learning-based CF methods and matching function learning-based CF methods. Representation learning-based CF methods try to map users and items into a common representation space. In this case, the higher similarity between a user and an item in that space implies they match better. Matching function learning-based CF methods try to directly learn the complex matching function that maps user-item pairs to matching scores. Although both methods are well developed, they suffer from two fundamental flaws, i.e., the limited expressiveness of dot product and the weakness in capturing low-rank relations respectively. To this end, we propose a general framework named DeepCF, short for Deep Collaborative Filtering, to combine the strengths of the two types of methods and overcome such flaws. Extensive experiments on four publicly available datasets demonstrate the effectiveness of the proposed DeepCF framework.

Gaussian-distributed codon frequencies of genomes

DNA encodes protein primary structure using 64 different codons to specify 20 different amino acids and a stop signal. Frequencies of codon occurrence when ordered in descending sequence provide a global characterization of a genome’s preference (bias) for using the different codons of the redundant genetic code. Whereas frequency/rank relations have been described by empirical relations, here we propose a statistical model in which two different forms of codon usage co-exist in a genome. We investigate whether such a model can account for the range of codon usages observed in a large set of genomes from different taxa. The differences in frequency/rank relations across these genomes can be expressed in a single parameter, the proportion of the two codon compartments. One compartment uses different codons with weak bias according to a Gaussian distribution of frequency, the other uses different codons with strong bias. In prokaryotic genomes both compartments appear to be present in a wide range of proportions, whereas in eukaryotic genomes the compartment with Gaussian distribution tends to dominate. Codon frequencies that are Gaussian-distributed suggest that many evolutionary conditions are involved in shaping weakly-biased codon usage, whereas strong bias in codon usage suggests dominance of few evolutionary conditions.

Rank relations between a {0, 1}-matrix and its complement

Let A be a {0, 1}-matrix and r(A) denotes its rank. The complement matrix of A is defined and denoted by Ac = J − A, where J is the matrix with each entry being 1. In particular, when A is a square {0, 1}-matrix with each diagonal entry being 0, another kind of complement matrix of A is defined and denoted by A = J − I − A, where I is the identity matrix. We determine the possible values of r(A) ± r(Ac) and r(A) ± r(A) in the general case and in the symmetric case. Our proof is constructive.

Competing through Co-operation in Nonhuman Primates: Developmental Aspects of Matrilineal Dominance

This paper's major aim is to illustrate how competition and co-operation are causally interrelated in the social life of nonhuman primates. In many species, competition for resources and mates commonly leads to xenophobic alliances and to the formation of intra-group dominance orders in which coalitions and alliances play a major role. In this sense, competition fosters co-operation. After briefly reviewing the nature of alliances in primates, this paper focuses on matrilineal dominance systems, which characterise many species of the Cercopithecidae family (macaques, baboons, etc.). In these societies, females socially inherit their mother's rank above lower-ranking matrilines with the result that kin rank next to each other. This paper summarises 10 years of experimental research on the composition and dynamics of alliances in a captive group of Japanese macaques (Macacafuscata). The main experimental paradigm consisted in manipulating the composition of the group, thereby dismantling existing alliances and inducing the formation of new alliances and new rank orders. Results reveal the existence of a complex interplay of kin and nonkin alliances responsible for the acquisition of rank and, later, for the maintenance of rank relations within and between kin groups. Opportunistic, selfishly motivated interventions in conflicts, performed by juveniles sharing the same targets (common targeting principle), appear to account for the initial formation of alliances in these dominance systems and, possibly, in various other situations as well. Such relatively simple processes may have paved the way for ontogenetically and phylogenetically more sophisticated forms of co-operation, such as reciprocity involving delayed benefits to the donors.