Search

The sparse triangular solver (SpTRSV) is mostly used for scientific and engineering applications. The structured-grid triangular solver of regular dependencies (STRSV) is a special kind of SpTRSV. Some general SpTRSVs that disregards the regularity of the matrix are unsuitable for solving this problem. This paper proposes an efficient parallel algorithm for STRSV on the SW26010 (a kind of China independently designed many-core processors), namely swStructTRSV. The algorithm makes full use of the fine-grained and low latency communication characteristics of the SW26010 to reduce the waiting time for synchronization, maximizes the regularity of access to improve memory access bandwidth, and achieves overlap between memory access and computation simultaneously. Moreover, the idea of the algorithm can be extended to incomplete LU factorization (ILU factorization) because of consistent dependencies. The experimental results on a core group(8 * 8 network composed of 64 cores) of SW26010 show that swStructTRSV can achieve an average speedup of over 30 in the sequential version. swStructTRSV on SW26010 achieves solving speedups of 2.2 and 6.3 over the fast STRSV (fSpTRSV) previously implemented on SW26010 and MKL on Intel Xeon Gold 6132, respectively. swStructTRSV significantly outperforms cuSparse on NVIDIA TITAN RTX in terms of overall execution time. [ABSTRACT FROM AUTHOR]

This paper presents a massively parallel approach of the multilevel fast multipole algorithm (PMLFMA) on homegrown many-core SW26010 cluster of China, noted as (SW-PMLFMA), for 3-D electromagnetic scattering problems. In this approach, the multilevel fast multipole algorithm (MLFMA) octree is first partitioned among management processing elements (MPEs) of SW26010 processors following the ternary partitioning scheme using the message passing interface (MPI). Then, the computationally intensive parts of the PMLFMA on each MPI process, matrix filling, aggregation and disaggregation are accelerated by using all the 64 computing processing elements (CPEs) in the same core group of the MPE via the Athread parallel programming model. Different parallelization strategies are designed for many-core accelerators to ensure a high computational throughput. In coincidence with the special characteristic of local Lagrange interpolation, the compressed sparse row (CSR) and the compressed sparse column (CSC) sparse matrix storage format is used for storing interpolation and anterpolation matrices, respectively, together with a specially designed cache mechanism of hybrid dynamic and static buffers using the scratchpad memory (SPM) to improve data access efficiency. Numerical results are included to demonstrate the efficiency and versatility of the proposed method. The proposed parallel scheme is shown to have excellent speedup. [ABSTRACT FROM AUTHOR]

In recent years, the convergence of the Internet of Things (IoT) and machine learning technologies has instigated advancements in IoT, and machine learning has revolutionized SMEs, reshaping their role in regional economic growth. These technologies provide SMEs with innovative tools, fostering connectivity within regional ecosystems. This transformation is crucial for China's economic realignment in its 'new normal' phase, with enterprises driving social and economic progress. This study proposed the XGBoost algorithm and IoT data enhance collaboration among SMEs in regional economies, fostering innovation and driving economic development. The role of IoT and machine learning in promoting sustainable regional economic prosperity. Technological innovation is essential for enterprise survival and growth in a competitive landscape. This paper examines the influence of environmental factors and geography on economic development across China's regions. It focuses on SME coupling and coordination, contributing to regional economic progress through performance evaluation. IoT integration provides SMEs with abundant real-time data, enabling deep insights into customer behavior, supply chains, and production. Simultaneously, the XGBoost algorithm efficiently processes and extracts actionable insights from this data. Data from 11 provinces in the Yangtze River economic belt highlights Jiangsu Province's leadership in regional innovation performance for 2015 and 2020. The empirical results, grounded in datasets amalgamating information from these provinces, attest to the potential of this IoT and XGBoost-driven approach. Boasting an impressive accuracy rate of 91.7%, surpassing alternative machine learning techniques like SVM, KNN, DT, RF, and LR, this research underscores the effectiveness of this integrated strategy in optimizing SME operations. Additionally, it computes the coupling coordination degree, innovation environment ranking, mean value, and numerical changes from 2010 to 2020, with an average innovation coupling coordination degree of 0.1624 across 30 Chinese provinces. [ABSTRACT FROM AUTHOR]

At present, the development of value chain mainly shows that the added value is positive; otherwise, the industrial value chain will be eliminated. Therefore, the construction and measurement of the industrial value chain height can effectively avoid the collapse of the industrial value chain caused by the low system value of the value chain. In this paper, the mobile virtualization channel allocation algorithm is used to construct the multi-objective industrial value chain height index, and the actual industrial value chain height in China is measured. This paper first describes the channel allocation algorithm of mobile virtualization, extracts the allocation rules of the algorithm, constructs the model of multi-objective industrial value chain, then measures the value chain of eight regions in China combined with the elements of technology, production, and marketing in the height of value chain, and quantitatively compares the results of value chain height of different regions according to the embeddedness and depth. Finally, it puts forward the improvement scheme for the measurement results and the problems faced by the value chain. The research shows that the average value chain height of South China is about 70.86, while the average value chain height of Northwest China is about 46.58, ranging from 40.58 to 60.45. Through quantitative data, the value chain height of South China is generally higher than that of Northwest China by more than 50%, which may be related to the innovation of technology and the imitation form of production. Therefore, this paper proposes to strengthen the technological innovation in the upstream of the value chain, so as to drive the development of the regional value chain. [ABSTRACT FROM AUTHOR]

This paper is a continuation of our previous paper [Front. Math. China, 2017, 12(5): 1023–1043] where global algorithms for computing the maximal eigenpair were introduced in a rather general setup. The efficiency of the global algorithms is improved in this paper in terms of a good use of power iteration and two quasi-symmetric techniques. Finally, the new algorithms are applied to Hua's economic optimization model. [ABSTRACT FROM AUTHOR]

In seismic data processing, low-frequency random noise with non-Gaussian and non-stationary characteristics heavily contaminates the reflected signals in Tarim area, which brings great difficulties in interpretation of seismic records in northwest China. To achieve more satisfied resolution, more greater fidelity, together with much higher increased signal-to-noise ratio (SNR), this paper proposes a SNR enhancement method based on the combination of variational mode decomposition (VMD) and Semi-soft Go Decomposition (Semi-Soft GoDec), named VMD-SSGoDec, which can realize the simplification of low-rank extraction in time–frequency representation (TFR) domain. Firstly, each trace of the rough seismic record is decomposed into several modes to reconstruct a component matrix by VMD. Due to the semi-low rank or approximate low-rank character of the desert low-frequency noise component matrix in TFR domain, secondly, we apply the Semi-soft GoDec, a low-rank matrix estimation to extract the low-frequency random noise components from the VMD results obtained in the first step. Repeating the above single-trace procedure to each trace rather than decomposing the entire record but use low-rank estimation once can lead to a more reduced dimension of the component matrix, and thus simplify the low-rank selection in Semi-soft GoDec. Finally, with the extracted random noise results in the second step, we can obtain the denoised record by making a difference with the original input. The proposed algorithm is tested by both synthetic record and field desert seismic data. Experimental results show outstanding advantages in low-frequency noise attenuation comparing with those of f-x deconvolution and SSWT-OptShrink. Both low-frequency random noise and surface waves are almost thoroughly attenuated by the proposed method, while the reflected signals are left nearly intact, revealing a significant enhancement in SNR. [ABSTRACT FROM AUTHOR]

Removing motion blur has been an important issue in computer vision literature. Motion blur is caused by the relative motion between the camera and the photographed object. However, in recent years, some achievements have been made in the research of image deblurring by using deep learning algorithms. In this paper, an enhanced adversarial network model is proposed. The proposed model can use the weight of feature channel to generate sharp image and eliminate draughtboard artefacts. In addition, the mixed loss function enables the network to output high-quality image. The proposed approach is tested using GOPRO datasets and Lai datasets. In the GOPRO datasets, the peak signal-to-noise ratio of the proposed approach is up to 28.674, and DeblurGAN is 27.454. And the structural similarity measure can be achieved up to 0.969, and DeblurGAN is 0.939. Furthermore, the images were obtained from China's Chang'e 3 Lander to test the new algorithm. Due to the elimination of the chessboard effect, the deblurred image has a better visual appearance. The proposed method achieved higher performance and efficiency in qualitative and quantitative aspects using the benchmark dataset experiments. The results also provided various insights into the design and development of the camera pointing system, which was mounted on the Lander for capturing images of the moon and rover for Chang'e space mission. [ABSTRACT FROM AUTHOR]

The disassembly line is the best way to deal with large-scale waste electrical and electronic equipment. Balancing of disassembly line is a hot and challenging problem in recent years. Given the uncertainty factors including corrosion and deformation of parts and components of waste products, this paper introduces the destructive mode and uncertainty disassembly time into the disassembly line and establishes a multi-objective disassembly line balancing model, considering partial destructive mode and U-shaped layout. The model aims to reduce the number of stations, balance the workload and reduce energy consumption while increasing the disassembly profit. A new multi-objective discrete flower pollination algorithm is proposed to solve the problem. Both task assignment and disassembly modes are considered in the encoding and decoding strategies of the flowers. Combining the discrete characteristics of the problem, the cross-pollination and self-pollination behaviors of the algorithm are redefined. The performance of the proposed algorithm is verified by solving two classical examples and by comparing with seven meta-heuristic algorithms. Then the proposed model and method are applied to a television disassembly line of a disassembly enterprise in China. The disassembly schemes of the proposed algorithm are superior to that of the five classical multi-objective algorithms. The results show that the proposed method can improve the performance of the disassembly line. [ABSTRACT FROM AUTHOR]

In order to realize the automatic epileptic seizure detection, feature extraction and classification of electroencephalogram (EEG) signals are performed on the interictal, the pre-ictal, and the ictal status of epilepsy patients. There is no effective strategy for selecting the number of channels and spatial filters in feature extraction of multichannel EEG data. Therefore, this paper combined sparse idea and greedy search algorithm to improve the feature extraction of common space pattern. The feature extraction can effectively overcome the repeating selection problem of feature pattern in the eigenvector space by the traditional method. Then we used the Fisher linear discriminant analysis to realize the classification. The results show that the proposed method can get high classification accuracy using fewer data. For 10 subjects, the averaged accuracy of epilepsy detection is more than 99%. So, the detection of an epileptic seizure based on sparse features using Fisher linear discriminant analysis classifiers is more suitable for a reliable, automatic epileptic seizure detection system to enhance the patient's care and the quality of life. [ABSTRACT FROM AUTHOR]

Logistics information will have a certain impact on the price of agricultural products. Therefore, the price of agricultural products can be analyzed and predicted based on the logistics information. Firstly, based on logistics information, a basic model of agricultural product price is constructed. Then, support vector machine prediction algorithm and ensemble learning prediction method are applied to analyze the price relationship of agricultural products and emotional characteristics are added. Secondly, the price data of all kinds of agricultural products in various provinces of China are collected, and the algorithms constructed in this paper are used to test the price data. The results show that the accuracy of each model is improved after adding emotional feature index, which indicates that emotional features can better supplement the shortcomings of digital features, thereby improving the accuracy of prediction. [ABSTRACT FROM AUTHOR]

Secondary use of medical big data is increasingly popular in healthcare services and clinical research. Understanding the logic behind medical big data demonstrates tendencies in hospital information technology and shows great significance for hospital information systems that are designing and expanding services. Big data has four characteristics - Volume, Variety, Velocity and Value (the 4 Vs) - that make traditional systems incapable of processing these data using standalones. Apache Hadoop MapReduce is a promising software framework for developing applications that process vast amounts of data in parallel with large clusters of commodity hardware in a reliable, fault-tolerant manner. With the Hadoop framework and MapReduce application program interface (API), we can more easily develop our own MapReduce applications to run on a Hadoop framework that can scale up from a single node to thousands of machines. This paper investigates a practical case of a Hadoop-based medical big data processing system. We developed this system to intelligently process medical big data and uncover some features of hospital information system user behaviors. This paper studies user behaviors regarding various data produced by different hospital information systems for daily work. In this paper, we also built a five-node Hadoop cluster to execute distributed MapReduce algorithms. Our distributed algorithms show promise in facilitating efficient data processing with medical big data in healthcare services and clinical research compared with single nodes. Additionally, with medical big data analytics, we can design our hospital information systems to be much more intelligent and easier to use by making personalized recommendations. [ABSTRACT FROM AUTHOR]

In brain MR images, the noise and low-contrast significantly deteriorate the segmentation results. In this paper, we propose an automatic unsupervised segmentation method integrating dual-tree complex wavelet transform (DT-CWT) with K-mean algorithm for brain MR image. Firstly, a multi-dimensional feature vector is constructed based on the intensity, the low-frequency subband of DT-CWT and spatial position information. Then, a spatial constrained K-mean algorithm is presented as the segmentation system. The proposed method is validated by extensive experiments using both simulated and real T1-weighted MR images, and compared with the state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

In this paper, we develop a novel feature selection algorithm based on the genetic algorithm (GA) using a specifically devised trace-based separability criterion. According to the scores of class separability and variable separability, this criterion measures the significance of feature subset, independent of any specific classification. In addition, a mutual information matrix between variables is used as features for classification, and no prior knowledge about the cardinality of feature subset is required. Experiments are performed by using a standard lung cancer dataset. The obtained solutions are verified with three different classifiers, including the support vector machine (SVM), the back-propagation neural network (BPNN), and the K-nearest neighbor (KNN), and compared with those obtained by the whole feature set, the F-score and the correlation-based feature selection methods. The comparison results show that the proposed intelligent system has a good diagnosis performance and can be used as a promising tool for lung cancer diagnosis. [ABSTRACT FROM AUTHOR]

Information fusion is a challenging task to extract the required information for high-level applications from various homogeneous or heterogeneous sensors. This paper summarizes some improvements in recent five years in the presented structure of information fusion, including estimation fusion, image fusion and others. Some technologies of information fusion and their variants are discussed. Although the state-of-the-art of the algorithms for information fusion have been proposed, there still remains some fundamental challenges with regard to exploiting the emerging multi-sensors' characteristics and their special structures. Finally, some potential prospects of estimation fusion and image fusion are discussed. [ABSTRACT FROM AUTHOR]

Since the late 1970s, special education in the People's Republic of China has experienced significant reform and fast development. However, education for children with severe developmental disabilities, especially autism spectrum disorders (ASDs), is still the greatest challenge in the field. This paper aims to give readers an overview of what is happening to children with ASDs in China. We first address the issue of prevalence of ASDs, and then offer an introduction to the diagnostic process. After that, a review of disability-related legislation is provided, followed by a description of current treatment options and available educational services. Finally we introduce all extent service providers and their roles. [ABSTRACT FROM AUTHOR]

A feature-weighted Support Vector Machine regression algorithm is introduced in this paper. We note that the classical SVM is based on the assumption that all the features of the sample points supply the same contribution to the target output value. However, this assumption is not always true in real problems. In the proposed new algorithm, we give different weight values to different features of the samples in order to improve the performance of SVM. In our algorithm, firstly, a measure named grey correlation degree is applied to evaluate the correlation between each feature and the target problem, and then the values of the grey correlation degree are used as weight values assigned to the features. The proposed method is tested on sample stock data sets selected from China Shenzhen A-share market. The result shows that the new version of SVM can improve the accuracy of the prediction. [ABSTRACT FROM AUTHOR]

In this paper, we apply finite-mixture-model-based clustering algorithms to cluster post-landfall tracks of tropical cyclones (TCs) making landfall over China. Because existing studies find that landfall surfaces or elevations affect post-landfall TC movements, we also take account of elevations in addition to time orders in this model. Our study reveals three clusters, with cluster 1 making landfall in Hainan province and moving across the western coast of Guangdong province. Most of the TC tracks in cluster 1 move over the ocean and make secondary landfalls over Yunnan province of China and Vietnam. Cluster 1 finally dissipates inland and moves westward as a result of the westward-shift subtropical high, westward steering flow, easterly vertical wind shear and strong mountainous blocking. Cluster 2 makes landfall over Guangdong and Fujian provinces. TCs in cluster 2 subsequently move inland and disappear due largely to westward-shift subtropical high, easterly steering flow, easterly vertical wind shear and relatively strong mountainous blocking. Cluster 3 makes landfall along the Fujian and Zhejiang coast and sustains a long period of time, recurving mostly to the mid-latitude region owing to the surrounding eastward-shift subtropical high, westerly vertical wind shear, weak mountainous blocking and westerly steering flow. Because cluster 2 is significantly associated with La Niña events, TCs more likely make landfall over southeastern China coast and move westward or northwestward without recurving. Cluster 3 sustains a longer time than clusters 1 and 2 in spite of its weak horizontal and vertical water vapor supply. TCs in cluster 3 interact actively with westerlies during the post-landfall period. However, we cannot observe any analogous interactions with the mid-latitude westerlies in clusters 1 and 2. TCs of clusters 1 and 2 are influenced by summer monsoon flows. Moreover, summer monsoon exerts a greater influence on cluster 1 than cluster 2. The composite 200 hPa divergence of cluster 3 is stronger than that of clusters 1 and 2. This explains to some degree why cluster 3 sustains longer than clusters 1 and 2 after making landfall. [ABSTRACT FROM AUTHOR]

In this paper, a time-varying AR model is constructed by using the vector-space algorithm of compactly-supported biorthonormal wavelet transform. It is developed for forecasting narrow monetary multipliers in China. [ABSTRACT FROM AUTHOR]