Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2022-05-10 Cooperative journals: 《计算机应用研究》
Abstract： This paper investigated the fault detection problem in finite-frequency domain for nonlinear networked systems under stochastic cyber-attacks. To save limited network resource, this paper introduced a novel dynamic event-triggered scheme. Firstly, under the consideration of fault sensitivity and disturbance robustness, the paper converted the addressed fault detection problem into an auxiliary #1; filtering problem by augmenting the states of the original system and the fault detection filter. Taking sector bounded nonlinearity and stochastic cyber-attacks into consideration, the design of #1; performance index included the frequency characteristics of fault signals. Combined with finite-frequency input characteristics, the paper proposed the joint design algorithm for fault detection filter and dynamic event-triggered scheme under the finite-frequency fault input. Finally, a simulation example of stirred tank reactor system verified the effectiveness of the proposed method.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2020-09-28 Cooperative journals: 《计算机应用研究》
Abstract： Based on convolution neural network, this paper proposed a new asymmetric two branch underwater biological classification model to solve the problems of low visibility, poor illumination conditions and no obvious differences among species in the underwater environment. In the model, the interactive branch used different convolution neural network to extracted local features and interacted with local features through the interactive module to enhanced the classification model. Convolutional neural network branch could effectively learned the global characteristics of the target and made up for the global information ignored in the interactive branch. Finally, this model obtains 98.9%, 98.3% and 97.9% of the accuracy on the three data sets of fish4 knowledge (f4k) , Eilat and RAMAS, which are significantly improved compared with the previous methods. visual interpretation also verifies that the model can effectively capture local features and eliminates the background influence. Finally, it shows that the model has good classification performance in underwater environment.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2019-05-10 Cooperative journals: 《计算机应用研究》
Abstract： Industrial processes often contain significant strong nonlinearity and time-varying behavior. Traditional extreme learning machine (ELM) based soft sensor sometimes fails to make use of data information effectively and has poor prediction performance. This paper proposed an improved particle swarm optimization algorithm. It has better convergence speed and search ability than standard particle swarm optimization (PSO) . This algorithm with an adaptive weight updates learning factor linearly and plays an important role in finding the optimal penalty coefficient and kernel parameter of ELM. The butane concentration of debutanizer column process simulation results verify that the proposed method has good prediction accuracy and generalization performance.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2019-01-28 Cooperative journals: 《计算机应用研究》
Abstract： In order to efficiently apply multi-label K-nearest neighbor(ML-KNN) in big dataset, this paper first conducted clustering algorithm to separate the training dataset into several parts, from which found a nearest cluster as new training set for test dataset; and then calculated nearest neighbor distance weight, by which the effect of k neighbors was described. Finally, an unseen sample could be classified by this new method. Numberical simulation results in different datesets show that a better classification result is obtained, and the time complexity of the algorithm is also improved.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-12-13 Cooperative journals: 《计算机应用研究》
Abstract： Aiming at the phenomenon of positive and negative correlation in the feature selection method of mutual information (MI) and the problem of not considering the word frequency of the feature items in different categories, a hybrid mutual information feature selection algorithm (HMI) is proposed. By introducing the inverse document frequency coefficient and the inter-class word frequency information coefficient, the algorithm can effectively utilize the word frequency information in the whole document and the word frequency information between each class. The positive and negative correlation coefficient is introduced to distinguish positive correlation and negative correlation and to make effective use. The experimental results show that the hybrid mutual information algorithm can effectively improve the quality of feature selection and then improve the effect of text emotional analysis.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-11-29 Cooperative journals: 《计算机应用研究》
Abstract： The spectral clustering is a clustering algorithm based on the theory of spectral partitioning, and it is a popular method due to its superior performance in the data sets with non-convex clusters. But the traditional spectral clustering algorithm cannot often get correct results on complex data sets, and the choice of parameters of affinity matrix construction depends on multiple tests and personal experience. Based on the situation, this paper proposes a spectral clustering algorithm based on natural nearest neighbor similarity graph（NSG-SC）. Natural nearest neighbor is a novel concept in terms of nearest neighbor, and it can avoid the disadvantages of K-nearest neighbor and ε-nearest neighbor. They usually need set parameters artificially effectively. The algorithm constructs an affinity matrix depending on the characteristics of the data sets, and it avoids some adverse effects. It is that inappropriate choice of parameters and isolated points cause them. The algorithm can also reflect better characteristics of data. The results of experiments show that the proposed algorithm named NSG-SC has feasibility and effectiveness.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-11-29 Cooperative journals: 《计算机应用研究》
Abstract： Path planning optimization problem has always been one of the important issues of the mobile robotics research. With the expansion of the fields of mobile robot applications and the complication of its working environment, traditional path planning algorithms become difficult to meet people's requirements due to their own limitations. In recent years, bioinspired intelligent algorithms（BIAs）are widely used in mobile robot path planning optimization because of its collective intelligence of the group and the characteristics of biological preference. First, this paper classifies the bioinspired intelligent algorithms used in the path planning optimization according to their sources of the biomimetic mechanism. Then, to optimize the latest research results and summarize existing problems and solutions, various new bioinspired intelligent algorithms used in the path planning optimization are described systematically according to different categories, and a detailed comparison and analysis of the algorithms performance in path planning optimization is also provided. Finally, further research direction about it will be discussed.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-10-11 Cooperative journals: 《计算机应用研究》
Abstract： In view of the problem that the overall time complexity of the FSDP clustering algorithm is high because the algorithm needs to traverse the entire data set when calculating the local density and minimum distance of data objects, this paper presents a Spark-based parallel FSDP clustering algorithm called SFSDP. First, the algorithm divides the dataset into multiple data partitions with relatively equal size by spatial meshing; then, the clustering analysis is performed on the data in each partition in parallel using the improved FSDP clustering algorithm; Global clusters are generated by grouping together local clusters between partitions. Experimental results show that SFSDP algorithm can effectively perform large-scale dataset clustering analysis compared with FSDP algorithm, and the algorithm has a good performance in terms of accuracy and scalability.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-05-24 Cooperative journals: 《计算机应用研究》
Abstract： Auto-Encoder can express the semantic features of data through deep unsupervised learning, but it is hard to determine the nodes of hidden layer and the processing of data for classification often leads to low accuracy and low stability. To solve the problems, this paper proposes a semi-supervised auto-encoder using sparse and label regularizations (LSRAE) to extract the essential features of the samples more accurately by combining unsupervised learning with supervised learning. The sparse regularization term adds constraints to the response of each hidden node, so that this algorithm can find potential structures in the data when the number of hidden neurons is large. At the same time, this algorithm introduce a label regularization term to compare the actual labels with desired labels by supervised learning to adjust the network parameters and further improve the classification accuracy. In order to verify the validity of the proposed method, this algorithm tests many data sets in the experiment. The results show that compared with traditional auto-encoders (AE) , sparse auto-encoder (SAE) , and extreme learning machine (ELM) , SLRAE can obviously improve the classification accuracy and stability when the processed data is applied to the same classifier.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-05-20 Cooperative journals: 《计算机应用研究》
Abstract： Due to the difference of handwriting caused by the large differences in edge contour, this paper proposed an algorithm named Jacobian regularized sparse automatic encoding machine (JSAE) for handwriting identification. This algorithm added sparse constraint and Jacobi regular item into the automatic coding machine, which improves the recognition accuracy. The sparse constraint can extract hidden structure from the data effectively and the regularized Jacobi can describe the marginal features of point data, thus it enables the learning ability of auto-encoder algorithm to improve and obtain the essential characteristics of the sample more accurately. Experimental results show that JSAE outperforms the basic auto-encoders (AE) and sparse auto-encoders(SAE) .
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-05-20 Cooperative journals: 《计算机应用研究》
Abstract： Abstract; In order to solve the problem that the histogram matching time is long and the engineering application is poor. This paper proposed to use EMD-L1 to measure the distance between two feature histograms. EMD-L1 fusion of the L1 norm and the original Earth Mover's Distance (EMD) and replace the calculation of the ground distance. The algorithm reduce the number of unknown variables, could achieve shape matching quickly and has a good retrieval performance. With a great deal of experiments in several shape databases, the results show that the performance of novel method are superior to original algorithm. And the matching speed is better than other algorithms under the MNIST data set.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-05-18 Cooperative journals: 《计算机应用研究》
Abstract： Researched on the task dividing method based on the architecture of the network emulation for the fusion of virtualization and digital simulation to improve the performance. This method took into account the advantages of virtualization and digital simulation, and the emulation network topology was divided into virtualization topology area and digital simulation topology area, and then divided the two topology area combined with given physical resources aiming at load balancing and remote traffic minimizing. Extensive experiments showed that using the method to divide the network emulation task, the remote traffic was reduced by 33.7%, 25.1% averagely, and the degree of load balancing was improved by 56.3%, 38.0% averagely, compared with the random algorithm and the uniform load balancing algorithm. The task dividing method can effectively reduce the remote traffic and improve the degree of load balancing.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-05-02 Cooperative journals: 《计算机应用研究》
Abstract： Fuzzy clustering is an important research content in the fields of pattern recognition, machine learning and image processing. Fuzzy C-means clustering algorithm is the most commonly used fuzzy clustering algorithm. The algorithm needs to preset the number of clusters in order to cluster the data set. This paper propose a new clustering validity index to validate the clustering results. This index defines the three important features of compactness, resolution and overlap degree from the perspective of partition entropy, membership degree and geometric structure. On this basis, this paper propose a method of determining the optimal clustering number. This paper validate the new clustering validity index and the traditional effectiveness index in six artificial data sets and three real data sets. The experimental results show that the proposed indexes and methods can effectively evaluate the clustering results and are suitable for determining the optimal clustering number of the samples.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-05-02 Cooperative journals: 《计算机应用研究》
Abstract： Aiming at the problem that the deep belief network (DBN) is susceptible to the training parameters燿uring the fine-tune process, this paper proposed a kind of batch normalization DBN classification method (BNDBN) . Firstly, this method used unsupervised learning to obtain high-level representation of raw data. Then through the introduction of scale transformation and translation transformation parameters, it processed the output characteristics of each layer by batch normalization. And it fed the post-processing characteristics into the nonlinear transformation activation layer. Finally, it trained and studied the parameters of the affine transformation and the original network by using the stochastic gradient descent method. The BNDBN method reduced the dependence of the gradient on the parameter size, which effectively resolved the problem of changing the value distribution of activation function caused by the change of network parameters and improves the training efficiency. To verify the effectiveness of the proposed method, it selected MNIST handwritten database and the USPS handwritten digital identification library for testing. Compared with the Dropout-DBN, DBN, ANN, SVM and KNN, the results show that the proposed method significantly improved the classification accuracy and had stronger feature extraction ability.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-04-19 Cooperative journals: 《计算机应用研究》
Abstract： Community structure is one of the important characteristics of Complex Network, the community detection algorithms with hierarchical clustering makes good use of modularity to achieve community detection, but its limitations lead to poor detection quality of networks which have complex structure of community and it may fail to identify communities smaller than a scale. This paper used local modularity to make up the lack of modularity on the basis of hierarchical clustering, it can avoid unreasonable result. The algorithm is tested on real-world data sets and artificial network, experimental results confirm that it has feasibility and effectiveness.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-04-17 Cooperative journals: 《计算机应用研究》
Abstract： Some research has introduced evolution idea into sampling algorithms, and related algorithms combined with adaptive Lévy distribution are proposed. This paper improves the evolutionary sampling algorithm based on Lévy distribution. By setting the parameter α of this distribution to 1.0, 1.3, 1.7, 2.0, corresponding to the four transition probability distributions, the diversity of the generated candidate samples is increased. Theoretical analysis and experimental results show that the proposed algorithm is superior to the evolutionary sampling algorithm based on Gaussian distribution, Cauchy distribution, symmetrical exponential distribution and other adaptive evolutionary sampling algorithms in terms of convergence rate and accuracy.
Hits
Hits
Downloads
Comment