System alignment supports cross-domain learning and zero-shot generalisation

Recent findings suggest conceptual relationships hold across modalities. For instance, if two concepts occur in similar linguistic contexts, they also likely occur in similar visual contexts. These similarity structures may provide a valuable signal for system alignment when learning to map between domains, such as when learning the names of objects. To assess this possibility, we conducted a paired-associate learning experiment in which participants mapped objects that varied on two visual features to locations that varied along two spatial dimensions. We manipulated whether the featural and spatial systems were \textit{aligned} or \textit{misaligned}. Although system alignment was not required to complete this supervised learning task, we found that participants learned more efficiently when systems aligned and that aligned systems facilitated zero-shot generalisation. We fit a variety of models to individuals' responses and found that models which included an offline unsupervised alignment mechanism best accounted for human performance. Our results provide empirical evidence that people align entire representation systems to accelerate learning, even when learning seemingly arbitrary associations between two domains.

Adaptive Collaborative Similarity Learning for Unsupervised Multi-view Feature Selection

In this paper, we investigate the research problem of unsupervised multi-view feature selection. Conventional solutions first simply combine multiple pre-constructed view-specific similarity structures into a collaborative similarity structure, and then perform the subsequent feature selection. These two processes are separate and independent. The collaborative similarity structure remains fixed during feature selection. Further, the simple undirected view combination may adversely reduce the reliability of the ultimate similarity structure for feature selection, as the view-specific similarity structures generally involve noises and outlying entries. To alleviate these problems, we propose an adaptive collaborative similarity learning (ACSL) for multi-view feature selection. We propose to dynamically learn the collaborative similarity structure, and further integrate it with the ultimate feature selection into a unified framework. Moreover, a reasonable rank constraint is devised to adaptively learn an ideal collaborative similarity structure with proper similarity combination weights and desirable neighbor assignment, both of which could positively facilitate the feature selection. An effective solution guaranteed with the proved convergence is derived to iteratively tackle the formulated optimization problem. Experiments demonstrate the superiority of the proposed approach.

SitNet: Discrete Similarity Transfer Network for Zero-shot Hashing

Hashing has been widely utilized for fast image retrieval recently. With semantic information as supervision, hashing approaches perform much better, especially when combined with deep convolution neural network(CNN). However, in practice, new concepts emerge every day, making collecting supervised information for re-training hashing model infeasible. In this paper, we propose a novel zero-shot hashing approach, called Discrete Similarity Transfer Network (SitNet), to preserve the semantic similarity between images from both ``seen'' concepts and new ``unseen'' concepts. Motivated by zero-shot learning, the semantic vectors of concepts are adopted to capture the similarity structures among classes, making the model trained with seen concepts generalize well for unseen ones benefiting from the transferability of the semantic vector space. We adopt a multi-task architecture to exploit the supervised information for seen concepts and the semantic vectors simultaneously. Moreover, a discrete hashing layer is integrated into the network for hashcode generating to avoid the information loss caused by real-value relaxation in training phase, which is a critical problem in existing works. Experiments on three benchmarks validate the superiority of SitNet to the state-of-the-arts.

Similarity and Group Performance

This paper investigates the effects that different patterns of similarity among group members have on a group's performance on a problem-solving task. I discuss and test hypotheses on the effects of similarity on group performance derived from two literatures: balance theory and research on homophily. In an experiment I found that the relative balance of the pattern of similarity was more important in predicting how quickly groups establish norms of interaction and complete a task than how similar group members were to each other. Neither balance nor the degree of similarity had a significant effect on the quality of the groups' work. I conclude that groups with balanced similarity structures produce task solutions that approximate the quality of those from other groups, but they do so in significantly less time. That is, balanced groups are more efficient than unbalanced groups.

Interesting Solvent Area in Crystal Structures of Two Natural Ergot Alkaloids

The structures of ergotamine bis(benzene) solvate (1) and ergocristine bis(benzene) solvate (2) are reported. Both structures crystallise in theP212121space group with cell parameters:1,a= 14.2968(3) Å,b= 15.4700(2) Å,c= 17.8123(4) Å, andV= 3939.57(13) Å3;2,a= 11.8358(2) Å,b= 17.6469(3) Å,c= 19.7125(3) Å, andV= 4117.25(12) Å3. Unexpectedly, despite the chemical similarity, structures of1and2significantly differ not only in the unit cell parameters, but also in the packing. Whereas in1solvent cavities are separated, there is only one unusual continuous solvent area in2filled with benzene, forming independent three-dimensional structure.

Cultural, Gender, and Individual Differences in Perceptual and Semantic Structures of Basic Colors in Chinese and English

In this paper we examine the judged similarity among the eight basic focal colors, and their names, among female and male Chinese (n = 68) and English (n = 52) speaking respondents. The major findings are: (1) all respondents share approximately sixty percent of their knowledge of the judged similarity structures of both semantic and perceptual tasks, (2) there are genuine individual differences among respondents that account for about fourteen percent of their knowledge on average, (3) there are small but statistically significant gender differences that come to about three percent on average, (4) there are small but statistically significant differences between Chinese and English respondents of about one-and-a-half percent, (5) there are differences in the semantic structure of the names of colors as compared to the judgments of the color samples that amounts to about five percent, and (6) there is about a three percent difference in the paired comparison task and the triads task. The results place strong constraints on theories relating to individual differences, linguistic relativity, and the relation of perceptual and semantic structures for colors.