Home      About Journal      Editorial Board      Instructions      Ethics Statement      Contact Us                   中文

Judgement and Disposal of Academic Misconduct Article Copyright Transfer Agreement Proof of Confidentiality Requirements for Electronic Version More.... Chinese Association of Automation National ResearchCenter for Intelligent Computing System Institute of Intelligent Machines,Chinese Academy of Sciences More....

2016 Vol.29 Issue.12, Published 2016-12-31 Papers and Reports    Researches and Applications        Papers and Reports 1057 Identification of Radiosensitivity Gene Signatures Based on Transcriptomic Data and Biological Network SHI Ming, WANG Hongqiang, SUN Tingting, XIE Xinping Accurate radiosensitivity prediction of tumor patients is crucial in tumor treatment. In this paper, a method is proposed for predicting tumor radiosensitivity based on high-throughput omics data in combination with biological networks. Firstly, Spearman correlation scores are calculated between gene expression profiles and tumor cell survival fraction at 2 Gy (SF2). Then, these scores are refined based on random walk theory by the topology of prior biological networks, such as gene networks or protein-to-protein interaction networks. Finally, highly significant radiosensitivity genes are screened out as feature variables to learn support vector machine(SVM) classifiers for predicting radiosensitivity of tumor patients. Experimental results on real-world microarrays datasets demonstrate the effectiveness of the proposed algorithm. 2016 Vol. 29 (12): 1057-1064 [Abstract] ( 655 ) [HTML 1KB] [ PDF 658KB] ( 546 ) 1065 Object Tracking via Global and Local Structural Inverse Sparse Appearance Model HU Zhengping, XIE Ronglu, WANG Meng, SUN Zhe To improve the performance of sparse representation based trackers, an object tracking method based on global and local structural inverse sparse appearance model is proposed. Firstly, an inverse sparse representation formulation is proposed to compute the weights of all particles by solving one optimization problem and this is conducive to improving the real-time performance. Then, a ranking mechanism based on joint discriminative similarity map(JDS map) is designed to improve the robustness. The formulation block candidates are divided into several pitches and the weighted sparse solutions are computed respectively. Next, these pitches are concatenated with different weights and meanwhile the sparse solution of each particle is computed. A combination mechanism is proposed to unite two sparse solutions as the JDS map. During the tracking, the object template and weight template are updated using a double-template updating strategy. Experiments demonstrate that the proposed algorithm is robust for benchmark video sequences under complicated conditions. 2016 Vol. 29 (12): 1065-1074 [Abstract] ( 454 ) [HTML 1KB] [ PDF 1586KB] ( 687 ) 1075 Chinese Microblog Sentiment Classification Based on Convolutional Neural Network LIAO Xiangwen, ZHANG Liyao, SONG Zhigang, CHENG Xueqi, CHEN Guolong To tackle the problems of the underutilization of context, the sparseness of data and the dependence on human-designed features in existing Chinese microblog sentiment classification methods, a Chinese microblog sentiment classification method based on convolutional neural network is proposed. Firstly, microblog messages are extended using the interaction context, and then they are initialized with dense vectors in the low-dimension space. Secondly, a convolutional neural network model is constructed for extracting and combining features. Finally, the sentiment of each microblog message is estimated by softmax function. Experimental results show that compared with baselines, the proposed method obtains higher accuracies and F1 values. 2016 Vol. 29 (12): 1075-1082 [Abstract] ( 694 ) [HTML 1KB] [ PDF 484KB] ( 981 ) 1083 Autonomous Exploration Approach of Mobile Robot Based on Visual FastSLAM CUI Shuai, GAO Jun, ZHANG Jun, FAN Zhiguo To meet the requirement of the indoor travelling and the localization of mobile robot, an approach for autonomous exploration, localization and mapping is proposed based on visual FastSLAM. Firstly, exploration position based on frontiers of the explored region is selected with consideration of information gain and path distance, and then path planning with the shortest distance to exploration position is performed to ensure the maximized exploration efficiency and completeness of the task accomplishment. FastSLAM 2.0 is employed as the basis of the proposed localization and mapping algorithm obtaining observation data by using robot vision, data observation efficiency is increased by fusing panoramic scanning and landmark tracking, and data association estimation is improved by introducing landmark visual information into calculation. The experimental results show that the proposed approach selects the best exploration position accurately, makes path planning reasonably, and accomplishes the exploration task successfully. The localization and mapping results of the proposed algorithm are robust with high accuracy. 2016 Vol. 29 (12): 1083-1094 [Abstract] ( 538 ) [HTML 1KB] [ PDF 1813KB] ( 1187 ) 1095 Comparative Study on Optimal Granularities in Inconsistent Multi-granular Labeled Decision Systems WU Weizhi, CHEN Chaojun, LI Tongjun, XU Youhong To study knowledge representation and knowledge acquisition in inconsistent decision systems with multi-granular labels, the concept of multi-granular labeled information systems is firstly introduced. Indiscernibility relations on the universe of discourse in a multi-granular labeled information system are defined. Representations of equivalence classes with different levels of granulation as well as their relationships are also explored. Lower and upper approximations of sets with different levels of granulation are further defined and their properties are presented. Finally, concepts of eight types of consistence and optimal granularity with various meanings in inconsistent multi-granular labeled decision systems are proposed and their relationships are examined. 2016 Vol. 29 (12): 1095-1103 [Abstract] ( 434 ) [HTML 1KB] [ PDF 380KB] ( 814 ) Researches and Applications 1104 Vehicle Moving Shadow Removal Approach Based on Zero-Tree Wavelet for Traffic Video WANG Xianghai, WANG Kai, LIU Meiyao, SU Yuanhe, SONG Chuanming The traditional background modeling method based on Gaussian model and the simple background subtraction method are difficult to accurately distinguish vehicles and shadows. Therefore， a vehicle moving shadow removal approach based on zero-tree wavelet (ZW) for traffic video is proposed in this paper. Firstly, the motion foreground image containing noise is converted to HSV color space and then the S channel and the V channel are processed with multilevel down-sampling wavelet transform. Secondly, by constructing the ZW mask of the motion foreground， the coefficients in different scale subbands are associated, and the mask values of fine scale subband can be guided and corrected by the father sub-band coefficients. Consequently, the accuracy of adaptive threshold of the subband is improved. By combining the shadow color characteristics， the distinction degree of judging vehicles and shadows is improved. A large number of simulation experiments verify the effectiveness of the proposed approach. 2016 Vol. 29 (12): 1104-113 [Abstract] ( 371 ) [HTML 1KB] [ PDF 1422KB] ( 504 ) 1114 Weighted Block Subspace Clustering Based on Least Square Regression LI Hui, CHEN Xiaoyun Traditional subspace clustering algorithms need to transform each sample into a vector form. Therefore, problems of high dimensionality and small size samples are caused, the natural structural information of each sample is ignored and the clustering information is missing. To overcome the drawbacks, the weighted block subspace clustering based on least square regression algorithm (WB-LSR) is proposed. Firstly, each sample is divided into lots of blocks, and the corresponding affinity matrices of each block are obtained. Next, the weight of each affinity matrix is determined by mutual vote between affinity matrices. Finally, the weighted sum of affinity matrices is regarded as final affinity matrix. The experimental results on image datasets and motion segmentation video datasets show that the proposed method effectively improves clustering accuracy. 2016 Vol. 29 (12): 1114-1121 [Abstract] ( 477 ) [HTML 1KB] [ PDF 407KB] ( 470 ) 1122 Air Quality Prediction Method Based on Fish Swarm and Fractal Dimension NI Zhiwei, ZHU Xuhui, CHENG Meiying To overcome defects of the existing air quality prediction method, an air quality prediction method based on fish swarm and fractal dimension is proposed. Firstly, the artificial fish are processed by discretization, the swarming and foraging behaviors and the moving way are improved, and the parallel mechanism and a strategy for overcoming local optimum are introduced. Secondly, air quality datasets are reduced by the improved discrete artificial fish swarm algorithm and the fractal dimension. Finally, an air quality prediction model is built by using Gaussian kernel SVM. Experiments are conducted on air quality datasets of Beijing, Shanghai and Guangzhou for nearly two years, and the experimental results show the relatively high stability and credibility of the proposed prediction method. 2016 Vol. 29 (12): 1122-1131 [Abstract] ( 436 ) [HTML 1KB] [ PDF 554KB] ( 740 ) 1132 Space Structure Based Affinity Propagation Algorithm for Categorical Data WANG Qi, QIAN Yuhua, LI Feijiang Constructing a reasonable similarity measure is difficult due to the lack of clear space structure in categorical data. Therefore, numerical clustering algorithms can hardly be extended to categorical data clustering. In this paper, a representation method for transforming the categorical data into numerical data is introduced. The similarity between samples is reconstructured and the structure feature of the original categorical data is maintained in the reconstruction process. Based on the data representation method, the affinity propagation(AP) clustering algorithm is migrated to the categorical data clustering. A space structure based AP algorithm for categorical data(SBAP) is proposed. Experimental results on several categorical datasets from the UCI dataset show that the proposed method makes AP algorithm deal with the categorical data clustering problem effectively with a significant improvement in performance. 2016 Vol. 29 (12): 1132-1139 [Abstract] ( 318 ) [HTML 1KB] [ PDF 392KB] ( 538 ) 1140 Consistency Based Partial Label Learning Algorithm TANG Caizhi, ZHANG Minling An essential strategy to solve the partial label problem is disambiguation. In most existing strategies, instances are individually disambiguated without the consideration of the relationships among instances. In this paper, a consistency based partial label learning (COPAL) algorithm is proposed assumpting that labels associated with similar instances are likely to be similar. Based on the above assumption, the labeling information of the instance itself and its neighboring instances are simultaneously utilized for disambiguation. Experiments on both artificial datasets and realworld datasets show the good generalization ability of COPAL.  2016 Vol. 29 (12): 1140-1146 [Abstract] ( 559 ) [HTML 1KB] [ PDF 446KB] ( 429 )

Supervised by China Association for Science and Technology Sponsored by Chinese Association of Automation NationalResearchCenter for Intelligent Computing System Institute of Intelligent Machines, Chinese Academy of Sciences Published by Science Press

Copyright © 2010 Editorial Office of Pattern Recognition and Artificial Intelligence Address: No.350 Shushanhu Road, Hefei, Anhui Province, P.R. China Tel: 0551-65591176 Fax:0551-65591176 Email: bjb@iim.ac.cn Supported by Beijing Magtech 　Email:support@magtech.com.cn