Quan Pan

Northwestern Polytechnical University, Xi’an, Liaoning, China

Are you Quan Pan?

Claim your profile

Publications (189)106.73 Total impact

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: A new fuzzy-belief K-nearest neighbor (FBK-NN) classifier is proposed based on evidential reasoning for dealing with uncertain data. In FBK-NN, each labeled sample is assigned with a fuzzy membership to each class according to its neighborhood. For each input object to classify, K basic belief assignments (bba's) are determined from the distances between the object and its K nearest neighbors taking into account the neighbors' memberships. The K bba's are fused by a new method and the fusion results are used to finally decide the class of the query object. FBK-NN method works with credal classification and discriminate specific classes, meta-classes and ignorant class. Meta-classes are defined by disjunction of several specific classes and they allow to well model the partial imprecision of classification of the objects. The introduction of meta-classes in the classification procedure reduces the misclassification errors. The ignorant class is employed for outliers detections. The effectiveness of FBK-NN is illustrated through several experiments with a comparative analysis with respect to other classical methods.
    Fusion 2014 Int Conf on Information Fusion, Salamanca, Spain; 07/2014
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The missing data in incomplete pattern can have different estimations, and the classification result of pattern with different estimations may be quite distinct. Such uncertainty (ambiguity) of classification is mainly caused by the loss of information in missing data. A new prototype-based credal classification (PCC) method is proposed to classify incomplete patterns using belief functions. The class prototypes obtained by the training data are respectively used to estimate the missing values. Typically, in a c-class problem, one has to deal with c prototypes which yields c estimations. The different edited patterns based on each possible estimation are then classified by a standard classifier and one can get c classification results for an incomplete pattern. Because all these classification results are potentially admissible, they are fused altogether to obtain the credal classification of the incomplete pattern. A new credal combination method is introduced for solving the classification problem, and it is able to characterize the inherent uncertainty due to the possible conflicting results delivered by the different estimations of missing data. The incomplete patterns that are hard to correctly classify will be reasonably committed to some proper meta-classes by PCC method in order to reduce the misclassification rate. The use and potential of PCC method is illustrated through several experiments with artificial and real data sets.
    Fusion 2014 Int Conf on Information Fusion, Salamanca, Spain; 07/2014
  • [Show abstract] [Hide abstract]
    ABSTRACT: The classification of incomplete patterns is a very challenging task because the object (incomplete pattern) with different possible estimations of missing values may yield distinct classification results. The uncertainty (ambiguity) of classification is mainly caused by the lack of information of the missing data. A new prototype-based credal classification (PCC) method is proposed to deal with incomplete patterns thanks to the belief function framework used classically in evidential reasoning approach. The class prototypes obtained by training samples are respectively used to estimate the missing values. Typically, in a c-class problem, one has to deal with c prototypes, which yield c estimations of the missing values. The different edited patterns based on each possible estimation are then classified by a standard classifier and we can get at most c distinct classification results for an incomplete pattern. Because all these distinct classification results are potentially admissible, we propose to combine them all together to obtain the final classification of the incomplete pattern. A new credal combination method is introduced for solving the classification problem, and it is able to characterize the inherent uncertainty due to the possible conflicting results delivered by different estimations of the missing values. The incomplete patterns that are very difficult to classify in a specific class will be reasonably and automatically committed to some proper meta-classes by PCC method in order to reduce errors. The effectiveness of PCC method has been tested through four experiments with artificial and real data sets.
    IEEE transactions on cybernetics. 07/2014;
  • [Show abstract] [Hide abstract]
    ABSTRACT: The classification of imprecise data is a difficult task in general because the different classes can partially overlap. Moreover, the available attributes used for the classification are often insufficient to make a precise discrimination of the objects in the overlapping zones. A credal partition (classification) based on belief functions has already been proposed in the literature for data clustering. It allows the objects to belong (with different masses of belief) not only to the specific classes, but also to the sets of classes called meta-classes which correspond to the disjunction of several specific classes. In this paper, we propose a new belief classification rule (BCR) for the credal classification of uncertain and imprecise data. This new BCR approach reduces the misclassification errors of the objects difficult to classify by the conventional methods thanks to the introduction of the meta-classes. The objects too far from the others are considered as outliers. The basic belief assignment (bba) of an object is computed from the Mahalanobis distance between the object and the center of each specific class. The credal classification of the object is finally obtained by the combination of these bba’s associated with the different classes. This approach offers a relatively low computational burden. Several experiments using both artificial and real data sets are presented at the end of this paper to evaluate and compare the performances of this BCR method with respect to other classification methods.
    Applied Intelligence 03/2014; · 1.85 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: We present a multidimensional evidential reasoning (MDER) approach to estimate change detection from the fusion of heterogeneous remote sensing images. MDER is based on a multidimensional (M-D) frame of discernment composed by the Cartesian product of the separate frames of discernment used for the classification of each image. Every element of the M-D frame is a basic joint state that allows to describe precisely the possible change occurrences between the heterogeneous images. Two kinds of rules of combination are proposed for working either with the free model, or with a constrained model depending on the integrity constraints one wants to take into account in the scenario under study. We show the potential interest of the MDER approach for detecting changes due to a flood in the Gloucester area in the U.K. from two real ERS and SPOT images.
    IEEE Geoscience and Remote Sensing Letters 01/2014; 11(1):168-172. · 1.82 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper we present a new credal classification rule (CCR) based on belief functions to deal with the uncertain data. CCR allows the objects to belong (with different masses of belief) not only to the specific classes, but also to the sets of classes called meta-classes which correspond to the disjunction of several specific classes. Each specific class is characterized by a class center (i.e. prototype), and consists of all the objects that are sufficiently close to the center. The belief of the assignment of a given object to classify with a specific class is determined from the Mahalanobis distance between the object and the center of the corresponding class. The meta-classes are used to capture the imprecision in the classification of the objects when they are difficult to correctly classify because of the poor quality of available attributes. The selection of meta-classes depends on the application and the context, and a measure of the degree of indistinguishability between classes is introduced. In this new CCR approach, the objects assigned to a meta-class should be close to the center of this meta-class having similar distances to all the involved specific classes׳ centers, and the objects too far from the others will be considered as outliers (noise). CCR provides robust credal classification results with a relatively low computational burden. Several experiments using both artificial and real data sets are presented at the end of this paper to evaluate and compare the performances of this CCR method with respect to other classification methods.
    Pattern Recognition. 01/2014; 47(7):2532–2541.
  • [Show abstract] [Hide abstract]
    ABSTRACT: Microbes play an important role on human health, however, little is known on microbes in the past decades for the limitation of culture-based techniques. Recently, with the development of next-generation sequencing (NGS) technologies, it is now possible to sequence millions of sequences directly from environments samples, and thus it supplies us a sight to probe the hidden world of microbial communities and detect the associations between microbes and diseases. In the present work, we proposed a supervised learning-based method to mine the relationship between microbes and periodontitis with 16S rRNA sequences. The jackknife accuracy is 94.83% and it indicated the method can effectively predict disease status. These findings not only expand our understanding of the association between microbes and diseases but also provide a potential approach for disease diagnosis and forensics.
    International Journal of Computational Biology and Drug Design 01/2014; 7(2/3):214-224.
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we present a new belief c×K neighbor (BCKN) classifier based on evidence theory for data classification when the available attribute information appears insufficient to correctly classify objects in specific classes. In BCKN, the query object is classified according to its K nearest neighbors in each class, and c×K neighbors are involved in the BCKN approach (c being the number of classes). BCKN works with the credal classification introduced in the belief function framework. It allows to commit, with different masses of belief, an object not only to a specific class, but also to a set of classes (called meta-class), or eventually to the ignorant class characterizing the outlier. The objects that lie in the overlapping zone of different classes cannot be reasonably committed to a particular class, and that is why such objects will be assigned to the associated meta-class defined by the union of these different classes. Such an approach allows to reduce the misclassification errors at the price of the detriment of the overall classification precision, which is usually preferable in some applications. The objects too far from the others will be naturally considered as outliers. The credal classification is interesting to explore the imprecision of class, and it can also provide a deeper insight into the data structure. The results of several experiments are given and analyzed to illustrate the potential of this new BCKN approach.
    Neurocomputing 01/2014; 133:459–470. · 1.63 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Particle filters provide a general numerical tool to deal with the nonlinear/non-Gaussian filtering problems. However, it is still a challenging problem to design a good proposal distribution to generate high-quality particles. In this paper, we present the concept of hybrid proposal distribution (HPD) defined by the weighted sum of multiple basic proposal distributions (BPDs), transform the adaptive particle filtering into the online weight optimization, and, as a result, propose the framework of particle filter with multimode sampling strategy. Compared with traditional sampling strategies, multimode sampling strategy is more flexible to accommodate the time-varying system characteristics. To demonstrate the efficiency of the proposed framework, a particle filter with HPD consisting of two BPDs is designed, where one BPD is the transition density and the other, first proposed in this paper, is defined by an updated system equation. The numerical simulation with two examples shows that the proposed filter outperforms the extended Kalman filter, the unscented Kalman filter, the standard particle filter and the unscented Kalman particle filter.
    Signal Processing. 11/2013; 93(11):3192–3201.
  • [Show abstract] [Hide abstract]
    ABSTRACT: This paper is motivated by the filtering estimation for a class of nonlinear stochastic systems in the case that the measurements are randomly delayed by one sampling time. Through presenting Gaussian approximation about the one-step posterior predictive probability density functions (PDFs) of the state and delayed measurement, a novel Gaussian approximation (GA) filter is derived, which recursively operates by analytical computation and Gaussian weighted integrals. The proposed GA filter gives a general and common framework since: (1) it is applicable for both linear and nonlinear systems, (2) by setting the delay probability as zero, it automatically reduces to the standard Gaussian filter without the randomly delayed measurements, and (3) many variations of the proposed GA filter can be developed through utilizing different numerical technologies for computing such Gaussian weighted integrals, including the previously existing EKF and UKF methods, as well as the improved cubature Kalman filter (CKF) in our paper using the spherical–radial cubature rule. The performance of the new method is demonstrated with a simulation example of the high-dimensional GPS/INS integrated navigation.
    IEEE Transactions on Automatic Control 04/2013; 49(4):976–986. · 2.72 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The K-nearest neighbor (K-NN) classification method originally developed in the probabilistic framework has serious difficulties to classify correctly the close data points (objects) originating from different classes. To cope with such difficult problem and make the classification result more robust to misclassification errors, we propose a new belief-based K-nearest neighbor (BK-NN) method that allows each object to belong both to the specific classes and to the sets of classes with different masses of belief. BK-NN is able to provide a hyper-credal classification on the specific classes, the rejection classes and the meta-classes as well. Thus, the objects hard to classify correctly are automatically committed to a meta-class or to a rejection class, which can reduce the misclassification errors. The basic belief assignment (bba) of each object is defined from the distance between the object and its neighbors and from the acceptance and rejection thresholds. The bba's are combined using a new combination method specially developed for the BK-NN. Several experiments based on simulated and real data sets have been carried out to evaluate the performances of the BK-NN method with respect to several classical K-NN approaches.
    Pattern Recognition. 03/2013; 46(3):834–844.
  • [Show abstract] [Hide abstract]
    ABSTRACT: The distributed node wakeup of wireless sensor networks is in the scope of collaborative optimization. Our recently-proposed artificial ant-colony (AAC) wakeup method for sensing modules (SMs) shows that the biologically-inspired idea is promising in significantly decreasing energy consumption while remaining the similar sensing performance, compared with the classical methods. However, the AAC method is hardly extended to the joint wakeup of SMs and communication modules (CMs) because the pheromone in the AAC cannot discern information from SMs or CMs. In other words, a novel biologically-inspired mechanism is needed. Inspired by the mechanism of disease propagation, a distributed infectious disease model (DIDM) is proposed including four sub-processes: direct infection, cross-infection immunity/immune deficiency, cross infection, and virus accumulation. Moreover, the DIDM based wakeup method is derived through establishing the correspondence between sensor wakeup and disease propagation. Besides, one theorem about parameter design is presented, exploiting the relationship among sensor properties, communication properties, performance requirements and the method parameters. The target-tracking simulation shows the effectiveness of our method.
    Information Sciences 01/2013; · 3.64 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The Probability Hypothesis Density (PHD) method can handle multi-target tracking problem, but it needs a specific association method to extract the target tracks. Up to now, such association methods are limited in the scope of temporal association, for example, the track labeling method. In this paper, we present the concept of the consistency measure between any two local peaks at the adjacent two time instants by using both spatial structure information and temporal evolution information. Furthermore, the global-space-time association is proposed through extracting the tracks one-by-one based on the consistency measure and three rules. The proposed method is testified via a simulation comparison with the track labeling method.
    Information Fusion (FUSION), 2013 16th International Conference on; 01/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: The evidential K-nearest neighbor (EK-NN) method, which extends the classical K-nearest neighbour (K-NN) rule within the framework of evidence theory, has achieved wide applications in pattern classification for its better performance. In EK-NN, the similarity of test samples with the stored training ones are assessed via the Euclidean distance function, which treats all attributes with equal importance. However, in many situations, certain attributes are more discriminative, while others may be less irrelevant, so attributes should be weighted differently in distance function. In this paper, a new evidential K-nearest neighbor classification method with weighted attributes (WEK-NN) is proposed to overcome the limitations of EK-NN. In WEK-NN, the class-conditional weighted Euclidean distance function is developed to assess the similarity between two objects and both a heuristic rule and a parameter optimization procedure are designed to derive the attribute weights. Several experiments based on simulated and real data sets have been carried out to evaluate the performance of the WEK-NN method with respect to several classical K-NN approaches.
    Information Fusion (FUSION), 2013 16th International Conference on; 01/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: Motivated by target tracking of OTHR inevitably facing the problem of multiple propagation modes, we present the problem of joint multipath data association and fusion (JMAF). Based on the expectation-maximum criterion and joint optimization, we derive the iterative optimization scheme of multipath data association and track fusion. A performance comparison of multipath probability data association (MPDA) is presented and discussed for a simple target tracking scenario, and the simulation results show that JMAF is superior to MPDA.
    Information Fusion (FUSION), 2013 16th International Conference on; 01/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: Matching-area suitability analysis in vision navigation system for unmanned aerial vehicle (UAV) is a very worthy but full of challenges research area. In this paper, a multi-feature fusion based visual saliency model (MFF-VSM) was established by introducing invariant features of speeded-up robust features (SURF) directly into the visual saliency model, based on which the extraction method of suitable matching-areas was proposed. With the integration of cross-scale SURF feature maps in the way we defined, the conspicuity map of SURF channel is obtained. By adding SURF channel into the traditional visual saliency model and fusing multi-feature of SURF, color, intensity and orientation, the MFF-VSM model is proposed. Based on the MFF-VSM, salient locations in sensed map could be obtained and chosen as suitable matching-areas. Simulation results show that the error of image registration with extracted matching-areas based on MFF-VSM meet the demands of vision navigation system. The proposed method may provide new ideas for autonomous navigation of UAV in the future.
    Information Fusion (FUSION), 2013 16th International Conference on; 01/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: The paper presents the problem of state estimation of linear stochastic time-varying system with generalized unknown disturbance (GUD) existing in the measurements. Such GUD can reflect the effects of sensor bias, deception jamming, navigation bias and so on. An upper-bound filter (UBF) is designed for such systems, and its optimal parameters are derived so that the minimum upper-bounds filter (MUBF) is obtained. The simulation about tracking a target via a biased sensor shows the effectiveness of the proposed filter.
    Information Fusion (FUSION), 2013 16th International Conference on; 01/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, the movement of a maneuvering low earth orbit satellite is modeled by a nonlinear stochastic system with unknown disturbance input, and an Iterated Minimum Upper Bound Filter is proposed to decrease the upper bound of the covariance of estimate errors via iterative optimization. The Monte Carlo simulation shows that the proposed filter significantly reduces the peak estimation errors due to orbit maneuvers compared with the well-known interacting multiple model method. Besides, it can accurately detect the target maneuvering time instant through thresholding the estimated fading factor.
    Information Fusion (FUSION), 2013 16th International Conference on; 01/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: Motivated by tracking a manoeuvring target in electronic counter environments, the authors present the problem of joint estimation and identification of a class of discrete-time stochastic systems with unknown inputs in both the plant and sensors. Based on the expectation-maximum criterion, the joint optimisation scheme of state estimation, parameter identification and iteration terminate decision were derived. A numerical example of tracking a manoeuvring target accompanied range gate pull-off is utilised to verify the proposed scheme.
    IET Control Theory and Applications 01/2013; 7(10):1377-1386. · 1.72 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Data clustering methods integrating information fusion techniques have been recently developed in the framework of belief functions. More precisely, the evidential version of fuzzy c-means (ECM) method has been proposed to deal with the clustering of proximity data based on an extension of the popular fuzzy c-means (FCM) clustering method. In fact ECM doesn’t perform very well for proximity data because it is based only on the distance between the object and the clusters’ center to determine the mass of belief of the object commitment. As a result, different clusters can overlap with close centers which is not very efficient for data clustering. To overcome this problem, we propose a new clustering method called belief functions cmeans (BFCM) in this work. In BFCM, both the distance between the object and the imprecise cluster’s center, and the distances between the object and the centers of the involved specific clusters for the mass determination are taken into account. The object will be considered belonging to a specific cluster if it is very close to this cluster’s center, or belonging to an imprecise cluster if it lies in the middle (overlapped zone) of some specific clusters, or belonging to the outlier cluster if it is too far from the data set. Pignistic probability can be applied for the hard decision making support in BFCM. Several examples are given to illustrate how BFCM works, and to show how it outperforms ECM and FCM for the proximity data.
    International Conference on Information Fusion (Fusion 2012), Singapore; 07/2012

Publication Stats

911 Citations
106.73 Total Impact Points

Institutions

  • 1997–2014
    • Northwestern Polytechnical University
      • • School of Automation
      • • Department of Control and Information Engineering
      Xi’an, Liaoning, China
  • 2010
    • University of Alberta
      • Department of Electrical and Computer Engineering
      Edmonton, Alberta, Canada
  • 2004–2010
    • Northwestern Polytechnic University
      China, Maine, United States
    • McMaster University
      • Department of Electrical and Computer Engineering
      Hamilton, Ontario, Canada
  • 2008
    • Shanghai Jiao Tong University
      • State Key Laboratory of Medical Genomics
      Shanghai, Shanghai Shi, China
    • Ruijin Hospital North
      Shanghai, Shanghai Shi, China
  • 2007
    • Nanjing Institute of Technology
      Nan-ching-hsü, Jiangxi Sheng, China
  • 2003–2006
    • Tsinghua University
      • Department of Automation
      Peping, Beijing, China
  • 2005
    • Robert-Bosch Krankenhaus
      Stuttgart, Baden-Württemberg, Germany