Loading...

Table of Content

    12 January 2017, Volume 0 Issue 1
    An Improved Collaborative Filtering Recommendation Algorithm
    LIU Yi, FENG Jun, WEI Tong-tong, CHEN Zhi-fei, XU Huan, ZHANG Li-xia
    2017, 0(1):  1-4+12.  doi:10.3969/j.issn.1006-2475.2017.01.001
    Asbtract ( 375 )  
    References | Related Articles | Metrics
    Recommendation system is widely used in e-commerce, and collaborative filtering is one of the most successful techniques in the recommendation system. With the increasing of the e-commerce data, the problem of the sparsity of the user-item rating matrix becomes more and more obvious, which has become the bottleneck of the recommendation system. To improve the recommendation quality under the sparse dataset environment, this paper proposed an improved collaborative filtering algorithm based on LDA model. We first built LDA model according to the user-item rating matrix, and got user-item selection probability matrix. And then, we clustered the item set by item properties, and cut the matrix by cluster results. Finally, in the process of similarity calculation, we introduced time factor to improve similarity calculation formula. Experimental results on Movie Lens datasets show that the proposed model gets better performance than traditional collaborative filtering algorithm in MAE.
    A Web News Extraction Method Based on Filtering Noise Wrapper
    SUN Meng, QU You-li
    2017, 0(1):  5-12.  doi:10.3969/j.issn.1006-2475.2017.01.002
    Asbtract ( 794 )  
    References | Related Articles | Metrics
    Extracting high-purity news from large Web pages, and stored in a structured form is the research foundation of the public opinion monitoring and topical updating. This paper proposes a Web news extraction method based on filtering noise wrapper. When inducing the wrapper, if the two strings don’t match each other, according to the threshold, we calculate the string tag path ratio of the strings to distinguish purity news from noise. At the same time, we propose two naive Bayes classifiers to extract the title and time of the news. Experimental results show that compared with other extraction technologies, the method in this paper has significant improvement in terms of accuracy and robustness. So it has greatly utility value.
    Short-term Wind Speed Forecasting Based on Local Gaussian Process
    CHANG Chun1, LI De-sheng2
    2017, 0(1):  13-16+22.  doi:10.3969/j.issn.1006-2475.2017.01.003
    Asbtract ( 199 )  
    References | Related Articles | Metrics
    Wind speed forecasting is very important to the operation of wind power plants and power system. A short-term wind speed forecasting method based on local Gaussian process model is proposed. Firstly, the training sample set is divided into many sub training set according to the fixed length of time window. Secondly, the local Gaussian process model is used to forecast the wind speed of each sub training set. By minimizing prediction error of the training set as the optimization goal, the improved PSO algorithm is used to optimize the hyper parameters. The prediction results show that the proposed method can improve the prediction accuracy.
    Swarm Intelligence Optimization Based on Estimation of Distribution Algorithm
    XUE Huan-ran1, JIANG Tao1, ZHENG Cong2
    2017, 0(1):  17-22.  doi:10.3969/j.issn.1006-2475.2017.01.004
    Asbtract ( 241 )  
    References | Related Articles | Metrics
    In order to solve the numerical optimization problem, the estimation of distribution algorithm based on normal distribution is researched and improved. On the basis of the existing algorithms, the replacement strategy with advantages, competing mechanism and pattern search are introduced. Compared with five kinds of improved algorithms, it is proved that the proposed strategies and the modified algorithms are effective and can find even better solutions. Finally, the effects of improvement are explained based on the numerical simulations and statistical analysis.
    Parameter-optimization of SVR Based on KFCM-MultiSwarmPSO
    MEN Hui-chao
    2017, 0(1):  23-31.  doi:10.3969/j.issn.1006-2475.2017.01.005
    Asbtract ( 218 )  
    References | Related Articles | Metrics
    Aiming at the various parameter-optimization methods of SVR, this paper proposes a new parameter-self-learning method, called MultiSwarmPSO, based on Multi-Swarm-PSO and KFCM to improve calculating efficiency, reduce the probability of mature convergence, and ameliorate the original algorithms. In this method, k-CV is blended in to improve the efficiency. Using power function as the dynamic learning factor to improve calculated performance is another innovative point in this paper. Specific to five different data sets, this paper compares the new method with grid algorithm, the standard PSO, the standard genetic algorithm, and artificial bee colony algorithm, the results show that this new method could improve the time efficiency and the fitting accuracy compared with the other two algorithms, and could determine the better parameters.
    An Ensemble Classification Algorithm Based on Frequent Subgraphs
    LIU Yi
    2017, 0(1):  32-35.  doi:10.3969/j.issn.1006-2475.2017.01.006
    Asbtract ( 230 )  
    References | Related Articles | Metrics
    Aiming at the contradiction of efficiency and correct rate existing in graph classification based on frequent subgraphs, the paper comes up with an algorithm for graph classification named G-Bagging. The algorithm makes base classifiers by traditional algorithm, and makes ensemble classifier by ensemble learning base classifiers, and updates ensemble classifier by redundancy management. Then we demonstrate that the algorithm can reduce the requirement of minimum support and training samples space by experiment, also is that the algorithm can ensure both efficiency and correct rate.
    Approach of Fault Diagnosis in Analog Circuit Based on Covering Algorithm
    DING Wei-cong, LI Zhi-hua, PEI Jie-cai
    2017, 0(1):  36-40.  doi:10.3969/j.issn.1006-2475.2017.01.007
    Asbtract ( 148 )  
    References | Related Articles | Metrics
    The research on theory and method of analog circuit fault diagnosis is a hot research topic at present. Traditional neural network learning algorithm has limitations, such as it is difficult to determine its structure, to comprehend and to achieve with the hardware. To solve these problems of traditional neural network in fault diagnosis of analog circuits, we used the neighborhood covering algorithm (NCA) to determine the structure of the neural network. In order to reduce number of neurons and determine a good initial point of NCA, this paper studied an improved method (NSCA). In the end, the fault diagnosis of a certain band-pass filter circuit, which reduces the number of neurons in the neural network, improves the generalization ability of the neural network, and improves the accuracy of diagnosis about 9 percentage point. Simulation results show that the method is more effective.
    Analog Circuit Fault Diagnosis Based on PSO-SVM of Hybrid Kernel Function
    PEI Jie-cai, LI Zhi-hua, DING Wei-cong
    2017, 0(1):  41-45+56.  doi:10.3969/j.issn.1006-2475.2017.01.008
    Asbtract ( 213 )  
    References | Related Articles | Metrics
    For the question caused by traditional support vector machine algorithm in analog circuit fault diagnosis, the way using support vector machine algorithm of hybrid kernel function (HSVM) and particle swarm optimization (PSO) is proposed. First, after analyzing the transient circuit under test, and writing down the output voltage, wavelet package is used to extract the output voltage feature; second, we use PSO to optimize the kernel weight and structure parameters of HSVM; last, the trained model is used to diagnose the fault. This method not only reduces the randomness of parameters selection, but also the accuracy of simulation result is improved 5%. The effectiveness is proved during the process of fault diagnosis in high-pass filter analog circuit.
    Transaction Design and Implementation for HDFS Access Middleware
    LI Na, CHEN Zheng-ming, LYU Jia, LIU Chun-fang
    2017, 0(1):  46-50.  doi:10.3969/j.issn.1006-2475.2017.01.009
    Asbtract ( 217 )  
    References | Related Articles | Metrics
    To settle the problem of Hadoop distributed file system (HDFS) not supporting transaction and difficult to recover from failure, a HDFS access middleware to support transaction for its application systems was proposed. Accessing to HDFS data, the solution can provide transaction supporting. Firstly, the transaction requirements for those applications systems accessing to HDFS were analyzed. Transaction process, transaction log and transaction recovery were described. Secondly, the transaction process and architecture of the middleware were designed. Thirdly, J2EE was used to implement the design of the proposed middleware. Finally, a test case was designed to test the proposed middleware design and implementation. It illustrates the proposed middleware can simplify both the process of accessing to HDFS and the process of recovering from failure.
    Relational Query in Large-scale Configuration Management Database Based on Graph Database
    DAI Sheng, WANG Bo
    2017, 0(1):  51-56.  doi:10.3969/j.issn.1006-2475.2017.01.010
    Asbtract ( 245 )  
    References | Related Articles | Metrics
    In a large-scale configuration management database (CMDB) which is built based on relationship databases, the function of relational query which is realized based on business scenes, turns out some disadvantages, it’s too complicated to construct a query and analysis statement, and it also needs a lot of time to execute. To solve these problems, the method to realize relational query function by using graph database is proposed. Making use of the consistency between the relationships of configuration items and graph data structure, the method constructs the configuration item relationships based on graph database. In order to realize the target of fast relational query, this paper designs and realizes a relational query module based on graph database, which is integrated into existing CMDB project with loose coupling. The experiment shows that the proposed method can solve the performance problem for relational query in CMDB.
    An Improved Dark Channel Prior Algorithm for Underwater Monitoring Image Processing
    KONG Qian-qian, WEN Ya-nan
    2017, 0(1):  57-60.  doi:10.3969/j.issn.1006-2475.2017.01.011
    Asbtract ( 193 )  
    References | Related Articles | Metrics
    With the development of marine fishery and underwater operation, under water monitoring and image processing play an important role in these areas. Due to the light attenuation of sea water and the scattering of suspended particles, the requiring of high resolution under water monitoring image is still a difficult work at present. This paper analyzes impact factors of underwater camera imaging quality and presents an improved dark channel prior algorithm for deep-sea camera imaging. The algorithm integrates the advantages of dark channel prior, image filtering and Retinex theory, implements defogging, noise reduction, color calibration as well as contrast enhancement. The experimental results show that the algorithm significantly improves the image sharpness and visual quality and can be widely used in deep-sea monitoring and image processing area.
    New Object Bank Method for Scene Classification Based on GBVS
    CHEN Meng-ting1, CHEN Si-xi2
    2017, 0(1):  61-64+70.  doi:10.3969/j.issn.1006-2475.2017.01.012
    Asbtract ( 214 )  
    References | Related Articles | Metrics
    Object bank (OB) representation is a novel image representation for high-level visual tasks, which encodes semantic and spatial information of the objects within an image. However, the poor precision of the object detectors in OB method influences the extraction effect of high-level image feature. In order to solve this problem, a new OB method improved by Graph-Based Visual Saliency (GBVS) is proposed. Firstly, GBVS saliency model is utilized to process the image and detect the saliency regions and extract better high-level feature. The experiments results show that the proposed method performs better in classification and increases the classification accuracy of 4%.
    A Review of Object Tracking Based on Mean Shift Algorithm
    LI Hui-xia1, LI Lin-sheng1, YAN Qing-sen2, ZHOU Jing-wen1
    2017, 0(1):  65-70.  doi:10.3969/j.issn.1006-2475.2017.01.013
    Asbtract ( 224 )  
    References | Related Articles | Metrics
    The mean shift algorithm and its research progress are introduced, and the video tracking method based on mean shift algorithm has been largely utilized in a wide-range of computer vision investigation and its practical application, especially in video tracking research. More importantly, among those existing object tracking algorithms, the mean shift algorithm could be able to solve numbers of critical problems during object tracking, such as sudden acceleration of the moving object, background interference, mutual occlusions among objects and/or between object and background, shape change of objects and/or background, etc. This paper describes the theory and applications based on improved mean shift algorithm and itself, including the details of those methods and their merits and demerits.
    Image Retrieval Based on Tensor Locality Preserving Projection
    HAN Dan-dan, HAN Li-xin
    2017, 0(1):  71-74.  doi:10.3969/j.issn.1006-2475.2017.01.014
    Asbtract ( 165 )  
    References | Related Articles | Metrics
    For locality preserving projections algorithm is unsupervised learning which is unable to distinguish the categories of information, this paper presents an improved locality preserving projection algorithm by adding discriminate information, and extends this algorithm to the tensor space. By comparing experimentally on standard data set, the retrieval results of the improved tensor locality preserving projection algorithm are more effective than the original algorithm.
    Pedestrian Detection and Tracking Under Dense Crowd Scene
    CAO Rui, WANG Min, DUAN Xiao-xiao
    2017, 0(1):  75-78.  doi:10.3969/j.issn.1006-2475.2017.01.015
    Asbtract ( 237 )  
    References | Related Articles | Metrics
    In this paper, we present a robust and efficient method to multiple humans tracking based on head detection. Methods of extracting background are failed due to the high density population and the serious problem of shade. The head in front of the video is detected by Viola and Jones AdaBoost cascade classifier based on Haar-like features, the back of the head is detected by Logistic regression based on head profile feature. After determining the pedestrian’s head position, the appearance color histograms for the head are modeled, finally we use a particle filter algorithm to track the pedestrian’s head. The experimental results demonstrate that the proposed method is capable of tracking humans effectively in high density crowds.
    Workshop Tool Management Subsystem
    GAO Peng-yu, LI Tao-ying,CHEN Yan
    2017, 0(1):  79-83.  doi:10.3969/j.issn.1006-2475.2017.01.016
    Asbtract ( 211 )  
    References | Related Articles | Metrics
    In order to better manage information and to create the greatest benefits with minimal cost, the development of specific information management system is particularly important to the specific functional requirements in specific enterprise. Based on workshop tool management subsystem in Dalian port, the article analyzes and designs monthly tool plan and amount of tool. In the Eclipse platform development environment, we apply the B/S structure based on MVC, use the WSH framework to complete the development of system. It makes the system function easy to implement and operate, and effectively improves the efficiency of management.
    Remote Diagnosis Data Support System for Electrical Equipments Based on Android Platform
    SHI Pan, GUO Ling
    2017, 0(1):  84-88+95.  doi:10.3969/j.issn.1006-2475.2017.01.017
    Asbtract ( 199 )  
    References | Related Articles | Metrics
    Remote diagnosis system for electrical equipments provides an efficient and low cost solution for fault diagnosis, but it relies on complete data about the equipments with fault. A data support system, as the subsystem of remote diagnosis platform, is presented, which is structured under the framework of SSH (Struts + Spring + Hibernate) architecture and jQuery mobile page scripts, and is developed with a comprehensive application of Eclipse, MySQL and Tomcat 8.0. The information about users, experts, companies, equipments, online measurement data, solutions, notice and users’ feedback are all managed in this system. The application shows the subsystem can provide powerful data support for the diagnosis process and has a high degree of integration, networking, intelligence, and is easy to maintain and expand.
    Design of Information Management System for Distributed Generation Monitoring
    SHENG Shen-yang, HE Xin
    2017, 0(1):  89-95.  doi:10.3969/j.issn.1006-2475.2017.01.018
    Asbtract ( 209 )  
    References | Related Articles | Metrics
    For the great randomness and weak controllability of distributed generation (DG), this paper designs a set of information management systems for distributed generation monitoring in order to realize the state monitoring, data acquisition, fault alarming, historic data management and trend analysis. The system is based on C/S architecture and uses the data compression methods such as dead zone compression and LZW lossless compression algorithms to build its own database system, which is composed of relational database and real time database. The system can acquire the current electricity information of DG (U/I/f/P/Q) and operation records (startup-stop, off-limits, etc), then analyze these information and store them to the database. The system can alarm with a variety of forms such as graphics, voice and cellphone.
    A Method of Security Risk Analysis for Network Based on Attack Detection and Vulnerability
    WANG Hong-kai1, YU Xiao-wen2, ZHANG Xu-dong1
    2017, 0(1):  96-100.  doi:10.3969/j.issn.1006-2475.2017.01.019
    Asbtract ( 169 )  
    References | Related Articles | Metrics
    With the development of Internet, passive defense gradually can not satisfy the requirement, active defense is more and more important. The base of active defense is security risk analysis of network, to analyze the risk, a method based on attack detection and vulnerability is put forward. By calculating the influence with attack information and nodes’ vulnerabilities, the risk of network can be determined and the leaks of the network system can be found. Compared with the method based on vulnerability, this method has a more perfect model and a more accurate result.
    Information Propagation Model in Online Social Network Based on PageRank
    CHEN Gao, WU Guang-chao
    2017, 0(1):  101-105.  doi:10.3969/j.issn.1006-2475.2017.01.020
    Asbtract ( 216 )  
    References | Related Articles | Metrics
    Classical information propagation models do not fully consider the complexity of online social networks and the differences of network topology structure between nodes. This paper proposed a new information propagation model in online social network based on PageRank (P-SIR). The model used PageRank of the node as a node authority and considered some transmission mechanisms in online social networks. It depicted the states evolution relationship between different types of nodes over time and reflected the news propagation process which was affected by the network topology structure and communication mechanism. The model also considered some actual factors in online social network which influenced the spreading of news. Using three different types of network to simulate the transmission process and analyze some impact factors, P-SIR model is verified by simulation experiment that it can effectively reflect the news propagation process of online social networks.
    Optimization Design and System Implementation of Flight Schedules Based on Least Cost
    FENG Xin1, XIA Hong-shan1, FENG Yue-gui2
    2017, 0(1):  106-110.  doi:10.3969/j.issn.1006-2475.2017.01.021
    Asbtract ( 209 )  
    References | Related Articles | Metrics
    To reduce the airline’s cost, this article designs an optimized mathematic model for flight schedule by considering eight key factors, such as the length of air route, total number of flights, number of passengers, intervals of flights, price of tickets, number of crewmen, personnel salary and fuel consumption. We solve the model with genetic algorithm and develop a visualized system with the Java platform. Experiments results show that the system is of good robustness and could reduce the operation cost of airlines and increase the profits effectively.
    Design of LED Plant Lighting System Based on Fuzzy Control
    CHEN Fang-yuan, TIAN Hong-xian
    2017, 0(1):  111-114+118.  doi:10.3969/j.issn.1006-2475.2017.01.022
    Asbtract ( 159 )  
    References | Related Articles | Metrics
    To solve the problem that the demand of plant growth lighting environment is nonlinear, time-varying and difficult to establish accurate mathematical model, this paper proposes a novel control system based on fuzzy control algorithm and fuzzy language description to realize the high precision control of plant growth lighting environment. In this system, PC interface based on LabVIEW, is used to set photon flux density (PFD) values of red, blue, green wavelength bands light for different plants or the same plant on different growth stages. In the lower computer system, 3 TEMT6000 light intensity sensors are used to detect the real-time PFD of red, blue and green bands and output the reasoning and decision-making according to voltage difference signal of red, blue, green light, then control quantity of LED light source module is achieved and PWM values are dynamically adjusted, thus intensity output of LED light in red, blue, green bands is controlled and PFD values of the plant lighting required for red, blue, green bands are ensured. Simulation and experiments show that the system realizes the intelligent adjustment to control red, blue, green light intensity output accurately and stablely, which meets the needs of plant growth lighting environment.
    Real-time Rendering Algorithm Optimization for Large Scale Transmission Lines Based on LOD
    DENG Yi-min1, TANG Zhi-qian2, LI Hong-bing3, YANG Zhong-ya4
    2017, 0(1):  115-118.  doi:10.3969/j.issn.1006-2475.2017.01.023
    Asbtract ( 203 )  
    References | Related Articles | Metrics
    A LOD algorithm by simulating the characteristics of the human eye and based on multiple control factors is proposed which is used to solve the problem of the low efficiency of the large-scale terrain environment simulation for transmission line corridor. The algorithm comprehensively considers the distance and terrain factors to simplify the terrain mesh, combined with the characteristics of the simulation scene mostly working in roaming mode. Moreover, in the evaluation function, the viewpoint movement speed is regarded as an important control factor. The optimized algorithm can simplify the calculation model of the evaluation function in the perspective of roaming state. The comparison experiment shows that the proposed algorithm has faster rendering speed than the ROAM algorithm and the restricted quadtree algorithm.
    Application and Challenge of Learning Analytics Based on Perspective of Users
    YUAN Fang, FU Da-jie, XU Wen
    2017, 0(1):  119-122.  doi:10.3969/j.issn.1006-2475.2017.01.024
    Asbtract ( 177 )  
    References | Related Articles | Metrics
    In the big data environment, with the rapid development of education information technology, learning analytics has been a hot research topic in recent years. In order to provide references for the follow-up study, based on the relevant literatures, this paper expounds the concept and characteristics of the learning analytics, from the user point of view of learners, teachers and teaching managers, and discusses the application of learning analytics in network learning and the problems and challenges faced by them.
    Simulation About Accurate Extraction of Emotional EEG
    TAN Zhi-wei, XIE Yun, SU Jing
    2017, 0(1):  123-126.  doi:10.3969/j.issn.1006-2475.2017.01.025
    Asbtract ( 196 )  
    References | Related Articles | Metrics
    EEG signal is a kind of microvolt signal. Collected from the scalp, EEG signals contain EOG signals, ECG signals as well as a variety of environmental noise. For the problem how the emotion recognition does effectively deal with EEG signals, firstly, the collected EEG signals from experiments are pretreated for removing interferences by using wavelet analysis and independent component analysis; secondly, in order to effectively extract EEG signal features, the amplitude histogram and standard difference are used to find qualitatively in time domain the differences in brain power of two kinds of emotion; lastly, the power spectrum is used to analyze the two emotional EEG gamma wave rhythm. The experimental simulation results show that it is feasible to use the gamma wave rhythm of EEG signals for emotion recognition.