Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2022-05-11 Cooperative journals: 《计算机应用研究》
Abstract： Aiming at the problems that the traditional offline hashing algorithm is time-consuming to train the model, occupies a large amount of memory and is difficult to update the model, and there is a large loss of labels in real image sets, this paper proposed a balanced label prediction for online hashing algorithm(BLPOH) . BLPOH generated predicted labels through the label prediction module, and fused the incomplete real labels, which can effectively alleviate the performance degradation of the model caused by label loss. It observed that there is an imbalance in the distribution of labels, then proposed a label category similarity balance algorithm to apply to the label prediction module to improve the accuracy of label prediction. It added the information of old data to the online update process of the hash function to improve the compatibility of the model to the old data. By conducting experiments on two widely used datasets and comparing with some state-of-the-art algorithms, the results demonstrate the superiority of BLPOH.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2022-05-10 Cooperative journals: 《计算机应用研究》
Abstract： Link prediction is an important issue in the subject of data mining. The similarity algorithm based on random walk often set the probability of particles transferring to adjacent nodes to be equal, but ignore the effect of node degree on the transition probability. To solve this problem, this paper proposed a link prediction algorithm based on lowest-degree preference random walk with restart. Firstly, the algorithm redefined the transition probability of the walkers by introduce lowest-degree preference function, then applied it to the random walk with restart, and explored the effect of lowest-degree preference strategy on node similarity. The experimental results of nine real networks show that the proposed method has higher prediction accuracy, and gives more network topology information, which proves that the algorithm has certain advantages in the evaluation of node similarity.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2020-09-28 Cooperative journals: 《计算机应用研究》
Abstract： This paper established a non-convex control model consists of time-invariant nonlinear system, non-affine equality constraint and non-convex inequality constrain aiming at the path planning problem of unmanned aerial vehicle (UAV) , along with an algorithm designed for solving the aforementioned model. Based on iterative optimization, it proposed the convex optimization iteration method and penalty function optimization strategy. The former uses the concave-convex process (CCCP) and Taylor formula to convexity the model, while the latter adds the processed term to the objective function as a penalty term to solve the feasibility limit of the initial point. It is proved that the proposed method strictly converges to a Karush-Kuhn-Tucker(KKT) point of the original problem. Simulation experiments result verified the feasibility and superiority of the penalty function convex optimization iteration algorithm, and it indicates that the proposed algorithm can provide a flight path satisfying the conditions for the unmanned aerial vehicle.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2019-04-01 Cooperative journals: 《计算机应用研究》
Abstract： Mixed integer nonlinear programming (MINLP) is widely used in scientific and engineering system designs. The traditional swarm intelligent algorithm fails to solve the contradiction between mining and exploration, competition and cooperation among individuals or populations in the population when solving mixed integer programming problems. In order to solve these two contradictions and optimize more efficiently, this paper proposed a PES algorithm due to swarm intelligent evolution strategy based on pyramid structure to solve the mixed integer programming problems. PES algorithm had a clear division of labor mechanism and promotion mechanism. A clear division of labor could balance global and local search capabilities. The promotion mechanism resolved the contradictions between competition and collaboration among populations. Standard test functions were used to simulate, and the experimental results showed that success rate and accuracy of the PES algorithm are better and verify effectiveness of the PES algorithm by comparing the results of the improved particle swarm optimization algorithm (CLSPSO, CLSPSO2) and the improved differential evolution algorithm (ridDE, ridDE2) .
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-05-24 Cooperative journals: 《计算机应用研究》
Abstract： Most existing Web image retrieval approaches only consider visual features. They ignore the valuable semantics involved in the associated texts, and fail to take advantages of text. This paper proposed a new unsupervised visual hashing approach called semantic transfer deep visual hashing (STDVH) . Firstly, it extracted the semantic information of the training text by spectral clustering. Then, it constructed a deep convolutional neural network to transfer the text semantic information into the learning of the image hash code. At last, it trained the image hash codes and hash functions in a unified framework, and completed the effective retrieval of large-scale image data in low-dimensional Hamming space. Experiments on two publicly available image datasets Wiki and MIR Flickr indicate that the proposed approach can achieve superior performance over other state-of-the-art techniques.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-05-18 Cooperative journals: 《计算机应用研究》
Abstract： Abstract : Overlapping community structure is one of the most important features of complex network. This study proposed a local extended genetic algorithm optimization overlapping community detection(LEGAOCD) . It makes a few core nodes be constructed as die body, yet regards the main idea of local extended overlapping community detection as reference; at the same time, this paper uses the triangular model to judge the stability measure of the community, so as to quantify the stability of community structure. Then, the improved strategy of genetic algorithm is used to allocate the communities where they belong. Finally, the high-quality overlapping community structure is obtained by two discriminant objective functions. After that, the LEGAOCD is compared with classical CPM and COPRA algorithms on the data sets, the results show that LEGAOCD possesses excellent comparatively in the aspects of detecting overlapping community structure and overlapping nodes.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-04-24 Cooperative journals: 《计算机应用研究》
Abstract： This paper introduced a discretization process to discretize fractional-order delay differential equations. First of all, it considered the fractional-order delay Wright fractional differential equation. Then, it analyzed the corresponding fractional-order Wright differential equation with piecewise constant arguments and applied the proposed discretization process to solve the model numerically. After that, according to the fixed points theory, this paper investigated the stability of the fixed points of the resultant dynamical system. Finally, it carried out a numerical simulation including Lyapunov exponent, phase diagrams, time series diagram, birurcation using Matlab to reveal more complex dynamics of the model. Simulation experiments show that this paper succeeds in discretizing fractional-order delay Wright equation.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-04-19 Cooperative journals: 《计算机应用研究》
Abstract： In reality, there are symmetry interdependent network and asymmetry interdependent network. At present, the cascading failures of symmetric interdependent networks are mainly studied. In order to make a comprehensive analysis of the robustness of the interdependent networks, this paper studied the cascading failure of asymmetric interdependent networks. Firstly, it constructed a cascaded failure model of asymmetric double layer interdependent networks using a typical BA scale-free network and a WS small world network. Then, under the condition of random attack and deliberate attack, it studied the dependence of network models, dependency patterns and the topology of sub networks on network failures. The simulation results show that the asymmetric dependent network is dependent regardless of the extent to which, in a random attack, positively related to degree of dependency networks exhibit strong robustness, while in deliberate attack, three different correlation shows the same robustness of interdependent networks. Moreover, the topology of the attack network affects cascading failures of the network.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-04-17 Cooperative journals: 《计算机应用研究》
Abstract： The prediction of the future income of highway toll has great guiding significance for the management and construction planning. However, the change of toll income is influenced by many factors. It has strong nonlinearity and complexity. The traditional prediction model cannot accurately express the development law of the toll income. In this paper, a highway toll prediction model based on gene expression programming algorithm (GEP) is established. The GEP algorithm is used to establish a complex functional relationship between current income and historical data, which accurately characterize the development rule of toll income over time. In addition, an effective correction model is proposed for the influence of toll reduction policies during holidays. Finally, this paper collects the historical data on the toll revenue of 12 companies such as shanghai-hangzhou-ningbo Expressway Co. , Ltd. Compared with traditional ARIMA and neural network prediction model, and the results fully verify the effectiveness and accuracy of the proposed algorithm.
Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-04-12 Cooperative journals: 《计算机应用研究》
Abstract： In order to use the high performance of fractional accumulated generating operator (FAGO) in the short-term grey prediction, in this paper, it firstly added FAGO in the time sequence variation grey model TSGM(1,1) to get higher accuracy. The main method organized as follow. Firstly, use the data of 1978 to 1987 from the monitoring station of “Lianziya” mountain in Hubei province as training data to optimal FAGO by using gravitational search algorithm and then use the data of 1988-1993 as verifying data to test the accuracy of the proposed grey model. Secondly, it compared other grey model GM(1,1) , fractional accumulated generating GM(1,1) and TSGM(1,1) . The result was that as follow. Firstly, corrected the error in the Chen's article. Moreover, it showed that the proposed model in this paper has higher prediction accuracy. Therefore, the novel model improves accuracy of the grey theory in long-term prediction by fractional accumulated generating operator and it provides the guide in the long-term prediction.
Hits
Hits
Downloads
Comment