Search

• Variational Quantum Circuits (VQCs) play a key role in several applications. • VQCs are parameterized quantum circuits trained by using classical optimizers. • The paper proposes to apply memetic algorithms to train VQCs. • The designed memetic algorithm outperforms the state-of-the-art classical optimizers. Among the ready-to-implement quantum algorithms, Variational Quantum Circuits (VQCs) play a key role in several applications, including machine learning. Their strength lies in the use of a parameterized quantum circuit that is trained by means of an optimization algorithm run on a classical computer. In such a scenario, there is a strong need to design appropriate classical optimization schemes that deal efficiently with VQCs and pave the way for quantum advantage in machine learning. Among possible optimization schemes, those based on evolutionary computation are finding increasing interest, given the unconventional and nonanalytical nature of the problem to be solved. This paper proposes to apply memetic algorithms to train VQCs used as quantum classifiers and shows the benefits of exploiting this evolutionary optimization technique through a comparative experimental session. [ABSTRACT FROM AUTHOR]

The Black Hole Clustering (BHC) algorithm is a density-based partitional clustering method inspired by the Density-based Spatial Clustering of Applications with Noise (DBSCAN). It does not require the number of clusters nor the computation of the pair-wise distance matrix between the data points, making it faster than DBSCAN. Also, it only needs one parameter that is intuitively easier to set than the epsilon parameter of DBSCAN. However, BHC needs the allocation of the so-called black holes that have to be linearly independent, making the algorithm in its current version suitable only for two or three-dimensional data sets. In this paper, we propose a generalized version of the black hole clustering algorithm (GBHC) by introducing a novel black hole allocation procedure for higher-dimensional data spaces. Furthermore, the proposed method is data-independent, so we have to run it once to obtain the black hole positions for all finite-dimensional metric spaces. We performed extensive computational experiments to compare GBHC with DBSCAN. The results show that both algorithms obtain comparable clustering solutions. GBHC, however, outperforms DBSCAN in computational complexity and explainability. • We propose a novel method to place the black holes for high-dimensional data sets. • This method is called Generalized Black Hole Clustering and it is data-independent. • We run the new method once to get black hole positions in all finite metric spaces. • We compare the GBHC and DBSCAN algorithms using several validation measures. [ABSTRACT FROM AUTHOR]

In view of the ability of the three-way concept lattices to describe the correlations and hierarchical relationships among the knowledge points, this paper proposes a knowledge point navigation approach based on the three-way concept lattices for autonomous learning. First, this paper constructs the formal context of the exercises-knowledge points according to the corresponding relationship between the exercises and knowledge points; then, the three-way concept lattice generation algorithms are used to obtain the AE-concept lattice and OE-concept lattice; finally, this paper provides students with an effective autonomous learning path and knowledge points navigation Hasse diagram to achieve efficient guidance of autonomous learning. [ABSTRACT FROM AUTHOR]

• The accuracy is greatly improved after replacing SCS module with the CSC module in Shift-GCN. • Inspired by shift CNNs, we replace shift space convolution in Shift-GCN with a sparse shift and named SparseShift-GCN. • The accuracy of algorithm was further improved after introduction of OHEM into SparseShift-GCN. Skeleton-based action recognition is widely used due to its advantages of lightweight and strong anti-interference. Recently, graph convolutional networks (GCNs) have been applied to action recognition and have made breakthrough progress. The shift convolution operator can effectively replace the spatial convolution and greatly reduce the computational complexity of the algorithm. This article first applies the Conv-Shift-Conv (CSC) module and the Shift-Conv-Shift-Conv (SC2) module to replace the Shift-Conv-Shift (SCS) module in spatial graph convolution of Shift-GCN respectively. This design can reorder the shifted channels more effectively. The experimental results show that the CSC module has the best effect and effectively improves accuracy of model. After that, this article proposes to replace the shift module in the original Shift-GCN with a sparse shift module and named SparseShift-GCN. This structure can reduce the redundancy of features, prevent overfitting and improve the generality of the model. Based on the improvement in the previous step, better results have been achieved. Finally, this paper uses OHEM Loss and Weighted Loss to carefully design the loss function of the model and introduces it into the model proposed in this paper. Experimental results show that OHEM Loss further improves the accuracy of algorithm. After a series of improvements, our proposed model has improved the accuracy of 4 different streams to varying degrees, which improves the overall performance of the network. [ABSTRACT FROM AUTHOR]

• Adaptively allocate computing resource to input regions for better network inference. • Patch sampling algorithm reduces redundant calculation costs in overlapping regions. • Comparable performance is achieved on two datasets by comparing with SOTA methods. [Display omitted] In this paper, we propose an entropy-dynamic resolution detection (EDRdet) method for object detection in aerial images. Most conventional object detection methods usually detect each region in aerial images directly with a fixed resolution, so that the resolution of the hard-to-detect region in the whole aerial image is not enough and that of the easy-to-detect region is not necessary. We argue that different resolutions of regions are required for efficient and accurate object detection in aerial images. Our EDRdet dynamically adjusts the resolution of each region via measuring the challenge of the detection process, therefore, our model can achieve a good trade-off between the computational cost and the accuracy effectively. Specifically, our EDRdet searches hard-to-detect regions through low-resolution global context information, and inputs higher-resolution patches to supplement lost feature at low resolution for fine-grained detection. We apply the entropy to determine the regions in aerial images required to be detected at higher resolutions, and where the entropy represents the uncertainty of the detection result. Moreover, we design a patch sampling algorithm to make the selected regions sparse to further improve the efficiency of patch generation. Extensive experiments on the DOTA and Visdrone2019 datasets verify that our EDRdet can reduce the computational cost and improve the model accuracy effectively. [ABSTRACT FROM AUTHOR]

• An efficient algorithm to update non-flat and incremental attributes in morphological trees. • Concepts and properties about updating non-flat and incremental attributes in morphological trees. • Adaptations of filter rules to compute additional structures for updating non-flat and incremental attributes. • Our algorithm is suitable for improving computation of composition of attribute filters [Display omitted] Attribute Filters are powerful image simplification operators that have very good contour-preservation properties. They can be efficiently computed using morphological trees. The most common strategy to compute attribute filters using morphological trees is based on three steps: (i) tree construction: the tree is built along with its attributes; (i i) filtering: we simplify the tree by removing some of its nodes based on some attribute and a filtering rule defined on a threshold value; and (i i i) image reconstruction: this is the final phase where the tree is converted back to an image leading to an attribute filter. However, after filtering a few attributes in tree may change (we call them as non-flat attributes) and they must be updated, when other attribute filters are applied to the simplified tree again, with either a different or the same threshold value. In this paper we present an efficient algorithm to update non-flat and incremental attributes in morphological trees with low memory consumption and fast computation. [ABSTRACT FROM AUTHOR]

In this paper, regularised regression for sequential data is investigated and a new ridge regression algorithm is proposed. It uses the Aggregating Algorithm (AA) to devise an iterative version of ridge regression (IRR). This algorithm is called AAIRR. A competitive analysis is conducted to show that the guarantee on the performance of AAIRR is better than that of the known online ridge regression algorithms. Moreover, an empirical study is carried out on real-world datasets to demonstrate the superior performance over those state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

• A lightweight CNN is designed for efficient gait recognition. • A joint knowledge distillation algorithm is proposed for boosting the performance of the simplified model. • Extensive experiments on two public datasets demonstrate the effectiveness of the proposed method. Gait recognition has made significant progress recently. However, most of existing methods utilize complicated neural networks, which lead to high computation cost. In this paper, a lightweight model named Distilled Light GaitSet (DLGS) is proposed for efficient gait recognition. More specifically, a lightweight CNN is designed for efficient computation, and a Joint Knowledge Distillation algorithm is proposed to boost the accuracy of the simplified model. Extensive experiments on the CASIA-B dataset and the OU-MVLP dataset show that the proposed DLGS can reduce the number of parameters and computation cost significantly while achieving the state-of-the-art performance. [ABSTRACT FROM AUTHOR]

• A novel PDE-constrained optimization problem is proposed for MRI denoising task. • The proposed optimization procedure computes the variable exponent and the clean image for MRI images. • The comparative results demonstrates the performance on two MRI datasets This paper investigates a novel PDE-constrained optimization model with discontinuous variable exponent p (x) identification. Since the parameter p is always related to a better approximation of the image gradient, its computation plays a critical role in preserving the image texture. Analytically, we include results on the approximation of this parameter as well as the resolution of the encountered PDE in a well posed framework. In addition, to resolve the PDE-constrained minimization problem, we proposed a modified primal-dual algorithm. Finally, numerical results are provided to compute the parameter p and also to remove high intensity of noise. The proposed algorithm simultaneously keep safe fine details and important features in medical image applications (Magnetic Resonance Images (MRI)) with numerous comparisons to show the performance of the proposed approach. [ABSTRACT FROM AUTHOR]

• Near optimal results, with a less than 5% deviation from optimal results of a DFS algorithm. • Fast execution time. • Scalable since it can deal with large data-sets in small execution time. • report tight upper-bounds for large data-sets. Graph Edit Distance (GED) problem is a well-known tool used to measure the similarity/dissimilarity between two graphs. It searches for the best set of edit operations (in terms of cost) that transforms one graph into another. Due to the NP-hardness nature of the problem, the search space increases exponentially making exact approaches impossible to use for large graphs. In this context, there is a huge need for approaches that give near-optimal results in reasonable time. In this paper, we propose a tree-based approximate approach for dealing with GED problem. It operates on a search tree that models all possible solutions of the problem. Since exploring the whole tree is impractical; this approach keeps only the best k nodes at each level of the tree for further exploration. This reduces enormously the execution time without scarifying the solution quality. Experiments using small and medium size data-sets show the low deviation of our results as compared to the optimal results of a Depth First Search algorithm. Moreover, our approach show a strong scalability potential by dealing with large data-sets in low execution time. [ABSTRACT FROM AUTHOR]

• A distributed algorithm is proposed to solve graph semi-supervised learning problem by leveraging the graph topology. • The convergence of the distributed algorithm is explicitly proved. • Numerical results verify the good performance and low cost of the distributed algorithm. Graph semi-supervised learning (GSSL) plays an important role in data classification by leveraging the similarity across the graph topology and convex optimization with Laplacian-based regularization. However, the current algorithm to solve the problem is centralized approach calling for heavy computational cost, particularly when the data is of large volume. In this paper, an innovate distributed algorithm is proposed to solve the problem, which is based on the decomposition of the similar graph. Contrary to the centralized approach, the distributed algorithm only requires the neighboring information for solving the optimization. It is proved that difference between the solutions of the distributed algorithm and centralized counterpart is upper bounded. We apply the proposed algorithm to both the synthetic and real-world datasets. The numerical results verify the effectiveness of the proposed distributed algorithm. [ABSTRACT FROM AUTHOR]

• A new branch is added to process structured data. • The objective function is modified to adapt to the new branch. • Detailed comparisons are made between our algorithm and the other classical models. The semi-supervised learning(SSL) was proposed to deal with the situation that only a few samples were labeled, so how to make the most use of the existing samples is crucial. In general, we apply data augmentation methods, like Generative Adversarial Networks(GAN), to increase the number of data when facing unstructured data. However, things get totally different with structured data, the continuity of neural networks limits their use in category variables. In this paper, we propose a semi-supervised Embedding GAN(EmGAN) to solve that problem. We add an embedding layer before the discriminator to better characterize category features and design a new loss function to further train our model. Moreover, the structures of the generator and discriminator are modified to match structured data. With the experiments of nonparametric statistical test, EmGAN shows its advantages in processing structured samples. [ABSTRACT FROM AUTHOR]

Many applications, such as robotics, are increasingly utilizing Reinforcement Learning (RL). The current implementation of RL typically assumes that the state update from the previous actions happens instantaneously, and thus can be used for making future decisions. However, this may not always be true. When the state update is not available, the decision is taken partly in the blind since it cannot rely on the current state information. This paper proposes an approach, where the delay in the knowledge of the state can be used, and the decisions are made based on the available information which may not include the current state information. One approach could be to include the actions after the last-known state as a part of the state information, however, that leads to an increased state-space making the problem complex and slower in convergence. The proposed algorithm gives an alternate approach where the state space is not enlarged, as compared to the case when there is no delay in the state update. Evaluations on the control environments further illustrate the improved performance of the proposed algorithm. [ABSTRACT FROM AUTHOR]

• The proposed method only needs a single discrete input parameter. • The proposed method has a robust performance in the different density scenario. • The proposed method can automatically detect the number of clusters. • We adapt the cluster stability for our method and find a new way to calculate it. The density-peak clustering (DPC) method (Rodriguez and Laio, 2014) clusters the data efficiently by fast searching density peaks. Recently, an improved DPC algorithm named 3DC method (Liang and Chen, 2016) was proposed for automatically detecting the correct structure of the clusters. However, it is difficult to select correct parameters for the DPC and 3DC methods and the local property of data set can't be revealed due to their global density assumption in some scenarios. To overcome this drawback, the K-nearest neighbor (KNN) framework is adapted for defining the density of the DPC method. Nevertheless, such KNN-based methods can't automatically detect the number of the clusters compared with the 3DC method. In this paper, an automatic clustering method is proposed, which needs only a discrete input parameter. Meanwhile, by utilizing the cluster stability for the Chameleon framework, the proposed method can automatically detect the correct structure of the clusters. The experimental results on the synthetic and real world data demonstrate that the proposed method has a more robust performance. Besides, the proposed method is robust to the choices of the input parameter. [ABSTRACT FROM AUTHOR]

• Optimizes the clustering result and the common representation subspace simultaneously across multiple views. • Untangle heavily mixed clusters by optimizing a single continuous objective. • Without the requirement of the number of clusters given in advance. • Not prone to stick into bad local minima even with outliers in data. This paper proposed a novel Robust Multi-View Continuous Subspace Clustering (RMVCSC) algorithm, which can untangle heavily mixed clusters by optimizing a single continuous objective. The proposed objective uses robust estimators to automatically clip specious inter-cluster connections while maintaining convincing intra-cluster correspondences in the common representation subspace learned from multiple views. The common representation subspace can reveal the underlying cluster structure in data. RMVCSC is optimized in an alternating minimization scheme, in which the clustering result and the common representation subspace are simultaneously optimized. Since different views can describe distinct perspectives of input data, the proposed algorithm has more accurate clustering performance than conventional algorithms by exploring information among multi-view data. In other words, the proposed algorithm optimizes a novel continuous objective in the simultaneously learned common representation subspace across multiple views. By using robust redescending estimators, the proposed algorithm is not prone to stick into bad local minima even with outliers in data. This kind of robust continuous clustering methods has never been used for multi-view clustering before. Moreover, the convergence of the proposed algorithm is theoretically proved, and the experimental results show that the proposed RMVCSC can outperform several very recent proposed algorithms in terms of clustering accuracy. [ABSTRACT FROM AUTHOR]

• We propose a DSNnet optimization model with non-convex equality constraints for unsupervised representation learning. • We design a neural network based algorithm to solve the proposed DSNnet model. • We introduce a novel similarity measurement between input data and learned features as a regularization term of DSNnet. Deep Semi-NMF (DSN), which learns hierarchical representations by stacking multiple layers Semi-NMF, shows competitive performance in unsupervised data analysis. However, the features learned from DSN always lack of representativity and discriminativity. In this paper, we build a novel Deep Semi-NMF network (DSNnet) to address the issues of DSN. Specifically, DSNnet contains multiple fully-connected layers, in which the activation function of each layer adopts Smoothly Clipped Absolute Deviation (SCAD). The non-negative hidden features are computed forwardly, while the network parameters are updated by the stochastic gradient descent method. Moreover, to enhance the discriminativity of features, we suggest simultaneously minimizing the reconstruction error of input and output, and maximizing the similarity between input and learned features. The proposed similarity measurement, which consists of global geometric similarity and local pointwise similarity, encourages the compactness between similar points and separateness between dissimilar points in the feature space, and is beneficial to preserve intrinsic information of original data. Extensive experiments conducted on several datasets illustrate the superiority of the proposed approach in comparison with state-of-the-art methods. [ABSTRACT FROM AUTHOR]

• A triple cross-domain collaborative filtering recommendation method is proposed. • An extended codebook construction algorithm is designed. • A transfer learning method based on extended codebook is presented. • Empirical results show the proposed method is superior to state-of-the-art methods. Recently, data sparsity is still one of the critical problems faced by recommendation systems. Although many existing methods based on cross-domain can alleviate it to a certain extent, these methods only use the information of single-domain (e.g., user-side, item-side and rating-side) or dual-domain (e.g., user-rating-side, user-item-side and item-rating-side) to make recommendations, which results in performance degradation. In this paper, we propose a triple cross-domain collaborative filtering method to alleviate data sparsity, named TCD-CF. In TCD-CF method, the triple-side intrinsic characteristics are first obtained by using the joint nonnegative matrix factorization to integrate the user-side, item-side and rating-side domain knowledge. Then the extended codebook (as knowledge to transfer) based on these intrinsic characteristics is constructed by using the orthogonal nonnegative matrix tri-factorization. Finally, the codebook-based transfer method for cross-system CF is applied into the source domain and target domain to predict the missing ratings and perform recommendation in the target domain. Extensive experiments on two real-world datasets demonstrate that the proposed method outperforms the state-of-the-art methods for the cross-domain recommendation task. [ABSTRACT FROM AUTHOR]

• Propose fast novel hard clustering algorithm for multi-type relational data. • Extend the popular NNDSVD method initialisation of our algorithm. • Propose internal performance clustering measure for assessing cluster similarity. Several cyber-security data sources are collected in enterprise networks providing relational information between different types of nodes in the network, namely computers, users and ports. This relational data can be expressed as adjacency matrices detailing inter-type relationships corresponding to relations between nodes of different types and intra-type relationships showing relationships between nodes of the same type. In this paper, we propose an extension of Non-Negative Matrix Tri-Factorisation (NMTF) to simultaneously cluster nodes based on their intra and inter-type relationships. Existing NMTF based clustering methods suffer from long computational times due to large matrix multiplications. In our approach, we enforce stricter cluster indicator constraints on the factor matrices to circumvent these issues. Additionally, to make our proposed approach less susceptible to variation in results due to random initialisation, we propose a novel initialisation procedure based on Non-Negative Double Singular Value Decomposition for multi-type relational clustering. Finally, a new performance measure suitable for assessing clustering performance on unlabelled multi-type relational data sets is presented. Our algorithm is assessed on both a simulated and real computer network against standard approaches showing its strong performance. [ABSTRACT FROM AUTHOR]

• Three-way concept stability and its properties. • Algorithm for the computation of three-way concept stability. • A case study on natural language generation scenario. Three-way concept analysis (3WCA) has been an emerging and important methodology for knowledge discovery and data analysis. Particularly, 3WCA can efficiently characterize the information of "jointly possessed" and "jointly not possessed" compared to the classical formal concept only can describe common attributes owned by objects. This property, typical of 3WCA has a huge potential in the field of Natural Language Generation (NLG). However, the construction of a three-way concept lattice is proved as an NP-complete problem and even harder than the construction of conventional concept lattice. This could negatively affect the use of 3WCA for NLG in real contexts. Hence, it is necessary to prune the three-way concept lattice and extract more interesting three-way concepts for knowledge acquisition. To this end, this paper defines the stability of a three-way concept and analyzes the relevant properties. An efficient computational algorithm for calculating the stability of three-way concepts is developed and evaluated by an experiment. In addition, a case study on NLG is conducted for demonstrating the applicability of the proposed technique. [ABSTRACT FROM AUTHOR]

• A scene image and human skeleton based dual-stream model is proposed. • A sparse sampling and video-level scene image classification method is adopted. • Spatio-temporal graph convolution is used for human skeleton feature extraction. • The proposed model has advantage of robustness to optical flow change. The dual stream-based human action recognition model offers the advantage of high recognition accuracy, but the algorithm is less robust in case of lighting changes. The human skeleton has a strong ability to express human behavior and actions; however, the scene information is ignored. Drawing on the idea of the dual-stream model, this paper proposes a human skeleton and scene image-based dual-stream model for human action recognition. The motion features are extracted through the spatio-temporal graph convolution of the human skeleton, and a scene recognition model is proposed based on the sparse frame sampling of video and video-level consensus strategy to process the scene video and gather the visual scene information. The proposed model exploits the advantages of skeleton information in motion expression and the superiority of the image in scene presentation. The scene information and spatio-temporal graph convolution-based human skeleton limbs are fused complementarily to achieve human action recognition. Compared to the conventional optical flow-based dual-stream action recognition method, this model is verified by experimenting under unstable light conditions, and the performance of human action recognition is robust and promising. [ABSTRACT FROM AUTHOR]

• A valid user modeling method which combines stable centroids and centroid fluctuation ranges. • A data cleaning algorithm in order to remove abnormal data. • A fixed centroid clustering algorithm by using the stable cluster centroids as the initial centroids. • An authentication method based on fluctuation ranges of centroid positions. User authentication of mobile devices is an important means to protect data security and user privacy. Low overhead, high accuracy and continuous authentication are challenging problems in user authentication. Aiming at the disadvantages of current authentication methods based on keystroke dynamics such as low accuracy and one-time authentication, this paper proposes a new authentication method named UIKI (User Identity authentication method based on clusters of Keystroke time Intervals). UIKI consists of the valid user modeling phase and the runtime user authentication phase. In the valid user modeling phase, UIKI uses a clustering algorithm to find the stable centroids and the fluctuation range of a valid user's centroid positions. In the runtime user authentication phase, UIKI clusters the data to be authenticated with the stable centroids as the initial centroids, and compares the results of clustering with the stable centroids and the centroid fluctuation range of the valid user to determine whether the input data are from the valid user. Experimental results prove that UIKI has the average FAR (False Accept Rate) of 0.082 and the average FRR (False Reject Rate) of 0.052, which can effectively authenticate a user's identity. UIKI has the advantages of high accuracy, low overhead and continuous authentication, which is suitable for the user authentication of mobile devices. [ABSTRACT FROM AUTHOR]