WeiGuo Zhang

Lianghao Xia, Chao Huang, Yong Xu et al.

As the deep learning techniques have expanded to real-world recommendation tasks, many deep neural network based Collaborative Filtering (CF) models have been developed to project user-item interactions into latent feature space, based on various neural architectures, such as multi-layer perceptron, auto-encoder and graph neural networks. However, the majority of existing collaborative filtering systems are not well designed to handle missing data. Particularly, in order to inject the negative signals in the training phase, these solutions largely rely on negative sampling from unobserved user-item interactions and simply treating them as negative instances, which brings the recommendation performance degradation. To address the issues, we develop a Collaborative Reflection-Augmented Autoencoder Network (CRANet), that is capable of exploring transferable knowledge from observed and unobserved user-item interactions. The network architecture of CRANet is formed of an integrative structure with a reflective receptor network and an information fusion autoencoder module, which endows our recommendation framework with the ability of encoding implicit user's pairwise preference on both interacted and non-interacted items. Additionally, a parametric regularization-based tied-weight scheme is designed to perform robust joint training of the two-stage CRANet model. We finally experimentally validate CRANet on four diverse benchmark datasets corresponding to two recommendation tasks, to show that debiasing the negative signals of user-item interactions improves the performance as compared to various state-of-the-art recommendation techniques. Our source code is available at https://github.com/akaxlh/CRANet.

Yan-Ping Wang, WeiGuo Zhang, Zhengbang Zha

Block ciphers use S-boxes to create confusion in the cryptosystems. Such S-boxes are functions over $\mathbb{F}_{2^{n}}$. These functions should have low differential uniformity, high nonlinearity, and high algebraic degree in order to resist differential attacks, linear attacks, and higher order differential attacks, respectively. In this paper, we construct new classes of differentially $4$ and $6$-uniform permutations by modifying the image of the Dobbertin APN function $x^{d}$ with $d=2^{4k}+2^{3k}+2^{2k}+2^{k}-1$ over a subfield of $\mathbb{F}_{2^{n}}$. Furthermore, the algebraic degree and the lower bound of the nonlinearity of the constructed functions are given.

Yunxiao Qin, Weiguo Zhang, Zezheng Wang et al.

Few-shot image classification (FSIC), which requires a model to recognize new categories via learning from few images of these categories, has attracted lots of attention. Recently, meta-learning based methods have been shown as a promising direction for FSIC. Commonly, they train a meta-learner (meta-learning model) to learn easy fine-tuning weight, and when solving an FSIC task, the meta-learner efficiently fine-tunes itself to a task-specific model by updating itself on few images of the task. In this paper, we propose a novel meta-learning based layer-wise adaptive updating (LWAU) method for FSIC. LWAU is inspired by an interesting finding that compared with common deep models, the meta-learner pays much more attention to update its top layer when learning from few images. According to this finding, we assume that the meta-learner may greatly prefer updating its top layer to updating its bottom layers for better FSIC performance. Therefore, in LWAU, the meta-learner is trained to learn not only the easy fine-tuning model but also its favorite layer-wise adaptive updating rule to improve its learning efficiency. Extensive experiments show that with the layer-wise adaptive updating rule, the proposed LWAU: 1) outperforms existing few-shot classification methods with a clear margin; 2) learns from few images more efficiently by at least 5 times than existing meta-learners when solving FSIC.

Yunxiao Qin, Weiguo Zhang, Chenxu Zhao et al.

Recently, meta-learning has been shown as a promising way to solve few-shot learning. In this paper, inspired by the human cognition process which utilizes both prior-knowledge and vision attention in learning new knowledge, we present a novel paradigm of meta-learning approach with three developments to introduce attention mechanism and prior-knowledge for meta-learning. In our approach, prior-knowledge is responsible for helping meta-learner expressing the input data into high-level representation space, and attention mechanism enables meta-learner focusing on key features of the data in the representation space. Compared with existing meta-learning approaches that pay little attention to prior-knowledge and vision attention, our approach alleviates the meta-learner's few-shot cognition burden. Furthermore, a Task-Over-Fitting (TOF) problem, which indicates that the meta-learner has poor generalization on different K-shot learning tasks, is discovered and we propose a Cross-Entropy across Tasks (CET) metric to model and solve the TOF problem. Extensive experiments demonstrate that we improve the meta-learner with state-of-the-art performance on several few-shot learning benchmarks, and at the same time the TOF problem can also be released greatly.

XiaoFang Wang, YuJuan Sun, WeiGuo Zhang

The state diagrams of a class of singular linear feedback shift registers (LFSR) are discussed. It is shown that the state diagrams of the given LFSR have special structures. An algorithm is presented to construct a new class of de Bruijn cycles from the state diagrams of these singular LFSR.

WeiGuo Zhang, ChunLei Xie, Enes Pasalic

In this paper, we employ the so-called semi-bent functions to achieve significant improvements over currently known methods regarding the number of orthogonal sequences per cell that can be assigned to a regular tessellation of hexagonal cells, typical for certain code-division multiple-access (CDMA) systems. Our initial design method generates a large family of orthogonal sets of sequences derived from vectorial semi-bent functions. A modification of the original approach is proposed to avoid a hard combinatorial problem of allocating several such orthogonal sets to a single cell of a regular hexagonal network, while preserving the orthogonality to adjacent cells. This modification increases the number of users per cell by starting from shorter codewords and then extending the length of these codewords to the desired length. The specification and assignment of these orthogonal sets to a regular tessellation of hexagonal cells have been solved regardless of the parity and size of $m$ (where $2^m$ is the length of the codewords). In particular, when the re-use distance is $D=4$ the number of users per cell is $2^{m-2}$ for almost all $m$, which is twice as many as can be obtained by the best known methods.