http://ieeexplore.ieee.org TOC Alert for Publication# 3477 2009 June 19 39 5 C1 1081 146 Nearest prototype methods can be quite successful on many pattern classification problems. In these methods, a collection of prototypes has to be found that accurately represents the input patterns. The classifier then assigns classes based on the nearest prototype in this collection. In this paper, we first use the standard particle swarm optimizer (PSO) algorithm to find those prototypes. Second, we present a new algorithm, called adaptive Michigan PSO (AMPSO) in order to reduce the dimension of the search space and provide more flexibility than the former in this application. AMPSO is based on a different approach to particle swarms as each particle in the swarm represents a single prototype in the solution. The swarm does not converge to a single solution; instead, each particle is a local classifier, and the whole swarm is taken as the solution to the problem. It uses modified PSO equations with both particle competition and cooperation and a dynamic neighborhood. As an additional feature, in AMPSO, the number of prototypes represented in the swarm is able to adapt to the problem, increasing as needed the number of prototypes and classes of the prototypes that make the solution to the problem. We compared the results of the standard PSO and AMPSO in several benchmark problems from the University of California, Irvine, data sets and find that AMPSO always found a better solution than the standard PSO. We also found that it was able to improve the results of the Nearest Neighbor classifiers, and it is also competitive with some of the algorithms most commonly used for classification. ]]> 39 5 1082 1091 243 In the past decades, many theoretical results related to the time complexity of evolutionary algorithms (EAs) on different problems are obtained. However, there is not any general and easy-to-apply approach designed particularly for population-based EAs on unimodal problems. In this paper, we first generalize the concept of the takeover time to EAs with mutation, then we utilize the generalized takeover time to obtain the mean first hitting time of EAs and, thus, propose a general approach for analyzing EAs on unimodal problems. As examples, we consider the so-called $(N + N)$ EAs and we show that, on two well-known unimodal problems, leadingones and onemax , the EAs with the bitwise mutation and two commonly used selection schemes both need $O(nln n + n^{2}/N)$ and $O(n lnln n + nln n/N)$ generations to find the global optimum, respectively. Except for the new results above, our approach can also be applied directly for obtaining results for some population-based EAs on some other unimodal problems. Moreover, we also discuss when the general approach is valid to provide us tight bounds of the mean first hitting times and when our approach should be combined with problem-specific knowledge to get the tight bounds. It is the first time a general idea for analyzing population-based EAs on unimodal problems is discussed theoretically. ]]> 39 5 1092 1106 290 A novel networked multisensor data-fusion method is developed in this paper. A federated filter is employed to fuse the data transmitted over the network, which plays an important role in the data-processing center. The stability of filters under the network is considered; an algorithm to deal with the delayed data is introduced, and the principle for data fusion is presented. Finally, two numerical examples show the effectiveness of the proposed scheme. ]]> 39 5 1107 1120 1216 In this paper, nonlinear system identification utilizing generalized total least squares (GTLS) methodologies in neurofuzzy systems is addressed. The problem involved with the estimation of the local model parameters of neurofuzzy networks is the presence of noise in measured data. When some or all input channels are subject to noise, the GTLS algorithm yields consistent parameter estimates. In addition to the estimation of the parameters, the main challenge in the design of these local model networks is the determination of the region of validity for the local models. The method presented in this paper is based on an expectation–maximization algorithm that uses a residual from the GTLS parameter estimation for proper partitioning. The performance of the resulting nonlinear model with local parameters estimated by weighted GTLS is a product both of the parameter estimation itself and the associated residual used for the partitioning process. The applicability and benefits of the proposed algorithm are demonstrated by means of illustrative examples and an automotive application. ]]> 39 5 1121 1133 606 A wireless sensor network (WSN) is a powerful unattended distributed measurement system, which is widely used in target surveillance because of its outstanding performance in distributed sensing and signal processing. This paper introduces a multiview visual-target-surveillance system in WSN, which can autonomously implement target classification and tracking with collaborative online learning and localization. The proposed system is a hybrid system of single-node and multinode fusion. It is constructed on a peer-to-peer (P2P)-based computing paradigm and consists of some simple but feasible methods for target detection and feature extraction. Importantly, a support-vector-machine-based semisupervised learning method is used to achieve online classifier learning with only unlabeled samples. To reduce the energy consumption and increase the accuracy, a novel progressive data-fusion paradigm is proposed for online learning and localization, where a feasible routing method is adopted to implement information transmission with the tradeoff between performance and cost. Experiment results verify that the proposed surveillance system is an effective, energy-efficient, and robust system for real-world application. Furthermore, the P2P-based progressive data-fusion paradigm can improve the energy efficiency and robustness of target surveillance. ]]> 39 5 1134 1146 729 Most existing active learning approaches are supervised. Supervised active learning has the following problems: inefficiency in dealing with the semantic gap between the distribution of samples in the feature space and their labels, lack of ability in selecting new samples that belong to new categories that have not yet appeared in the training samples, and lack of adaptability to changes in the semantic interpretation of sample categories. To tackle these problems, we propose an unsupervised active learning framework based on hierarchical graph-theoretic clustering. In the framework, two promising graph-theoretic clustering algorithms, namely, dominant-set clustering and spectral clustering, are combined in a hierarchical fashion. Our framework has some advantages, such as ease of implementation, flexibility in architecture, and adaptability to changes in the labeling. Evaluations on data sets for network intrusion detection, image classification, and video classification have demonstrated that our active learning framework can effectively reduce the workload of manual classification while maintaining a high accuracy of automatic classification. It is shown that, overall, our framework outperforms the support-vector-machine-based supervised active learning, particularly in terms of dealing much more efficiently with new samples whose categories have not yet appeared in the training samples. ]]> 39 5 1147 1161 806 A novel reinforcement-learning-based output adaptive neural network (NN) controller, which is also referred to as the adaptive-critic NN controller, is developed to deliver the desired tracking performance for a class of nonlinear discrete-time systems expressed in nonstrict feedback form in the presence of bounded and unknown disturbances. The adaptive-critic NN controller consists of an observer, a critic, and two action NNs. The observer estimates the states and output, and the two action NNs provide virtual and actual control inputs to the nonlinear discrete-time system. The critic approximates a certain strategic utility function, and the action NNs minimize the strategic utility function and control inputs. All NN weights adapt online toward minimization of a performance index, utilizing the gradient-descent-based rule, in contrast with iteration-based adaptive-critic schemes. Lyapunov functions are used to show the stability of the closed-loop tracking error, weights, and observer estimates. Separation and certainty equivalence principles, persistency of excitation condition, and linearity in the unknown parameter assumption are not needed. Experimental results on a spark ignition (SI) engine operating lean at an equivalence ratio of 0.75 show a significant (25%) reduction in cyclic dispersion in heat release with control, while the average fuel input changes by less than 1% compared with the uncontrolled case. Consequently, oxides of nitrogen $(hbox{NO}_{x})$ drop by 30%, and unburned hydrocarbons drop by 16% with control. Overall, $hbox{NO}_{x}$'s are reduced by over 80% compared with stoichiometric levels. ]]> 39 5 1162 1179 1131 An adaptive output feedback neural network controller is designed, which is capable of rendering affine-in-the-control uncertain multi-input–multi-output nonlinear systems strictly passive with respect to an appropriately defined set. Consequently, a simple output feedback is employed to stabilize the system. The controlled system need not be in normal form or have a well-defined relative degree. Without requiring a zero-state detectability assumption, uniform ultimate boundedness, with respect to an arbitrarily small set, of both the system's state and the output is guaranteed, along with boundedness of all other signals in the closed loop. To effectively avoid possible division by zero, the proposed adaptive controller is of switching type. However, its continuity is guaranteed, thus alleviating drawbacks connected to existence of solutions and chattering phenomena. Simulations illustrate the approach. ]]> 39 5 1180 1191 248 We consider the microaggregation problem (MAP) that involves partitioning a set of individual records in a microdata file into a number of mutually exclusive and exhaustive groups. This problem, which seeks for the best partition of the microdata file, is known to be NP-hard and has been tackled using many heuristic solutions. In this paper, we present the first reported fixed-structure-stochastic-automata-based solution to this problem. The newly proposed method leads to a lower value of the information loss (IL), obtains a better tradeoff between the IL and the disclosure risk (DR) when compared with state-of-the-art methods, and leads to a superior value of the scoring index, which is a criterion involving a combination of the IL and the DR. The scheme has been implemented, tested, and evaluated for different real-life and simulated data sets. The results clearly demonstrate the applicability of learning automata to the MAP and its ability to yield a solution that obtains the best tradeoff between IL and DR when compared with the state of the art. ]]> 39 5 1192 1205 428 We propose two variations of the support vector data description (SVDD) with negative samples (NSVDD) that learn a closed spherically shaped boundary around a set of samples in the target class by involving different forms of slack vectors, including the two-norm NSVDD and $nu$-NSVDD. We extend the NSVDDs to solve the multiclass classification problems based on the distances between the samples and the centers of the learned spherically shaped boundaries in a kernel-defined feature space by using a combination of linear discriminant analysis (LDA) and nearest-neighbor (NN) rule. Extensive simulations are developed with one real-world data set on the automatic monitoring of roller bearings with vibration signals and eight benchmark data sets for both binary and multiclass classification. The benchmark testing results show that our proposed methods provide lower classification error rates and smaller standard deviations with the cross-validation procedure. The two-norm NSVDD with the LDA–NN rule recorded a test accuracy of 100.0% for the binary fault detection of roller bearings and 99.9% for the multiclass classification of roller bearings under six conditions. ]]> 39 5 1206 1216 1312 In many current face-recognition (FR) applications, such as video surveillance security and content annotation in a web environment, low-resolution faces are commonly encountered and negatively impact on reliable recognition performance. In particular, the recognition accuracy of current intensity-based FR systems can significantly drop off if the resolution of facial images is smaller than a certain level (e.g., less than 20 $times$ 20 pixels). To cope with low-resolution faces, we demonstrate that facial color cue can significantly improve recognition performance compared with intensity-based features. The contribution of this paper is twofold. First, a new metric called “variation ratio gain” (VRG) is proposed to prove theoretically the significance of color effect on low-resolution faces within well-known subspace FR frameworks; VRG quantitatively characterizes how color features affect the recognition performance with respect to changes in face resolution. Second, we conduct extensive performance evaluation studies to show the effectiveness of color on low-resolution faces. In particular, more than 3000 color facial images of 341 subjects, which are collected from three standard face databases, are used to perform the comparative studies of color effect on face resolutions to be possibly confronted in real-world FR systems. The effectiveness of color on low-resolution faces has successfully been tested on three representative subspace FR methods, including the eigenfaces, the fisherfaces, and the Bayesian. Experimental results show that color features decrease the recognition error rate by at least an order of magnitude over intensity-driven features when low-resolution faces (25 $times$ 25 pixels or less) are applied to three FR methods. ]]> 39 5 1217 1230 1167 The main objective of this paper is to present a new optimization approach, which we call heuristic Kalman algorithm (HKA). We propose it as a viable approach for solving continuous nonconvex optimization problems. The principle of the proposed approach is to consider explicitly the optimization problem as a measurement process designed to produce an estimate of the optimum. A specific procedure, based on the Kalman method, was developed to improve the quality of the estimate obtained through the measurement process. The efficiency of HKA is evaluated in detail through several nonconvex test problems, both in the unconstrained and constrained cases. The results are then compared to those obtained via other metaheuristics. These various numerical experiments show that the HKA has very interesting potentialities for solving nonconvex optimization problems, notably concerning the computation time and the success ratio. ]]> 39 5 1231 1244 448 This paper is concerned with the problem of designing fuzzy controllers for a class of nonlinear dynamic systems. The considered nonlinear systems are described by T-S fuzzy models with nonlinear local models, and the fuzzy models have fewer fuzzy rules than conventional T-S fuzzy models with local linear models. A new fuzzy control scheme with local nonlinear feedbacks is proposed, and the corresponding control synthesis conditions are given in terms of solutions to a set of linear matrix inequalities (LMIs). In contrast to the existing methods for fuzzy control synthesis, the new proposed control design method is based on fewer fuzzy rules and less computational burden. Moreover, the local nonlinear feedback laws in the new fuzzy controllers are also helpful in achieving good control effects. Numerical examples are given to illustrate the effectiveness of the proposed method. ]]> 39 5 1245 1258 569 In this paper, we address the problem of fusing information about object positions in multirobot systems. Our approach is novel in two main respects. First, it addresses the multirobot object localization problem using fuzzy logic. It uses fuzzy sets to represent uncertain position information and fuzzy intersection to fuse this information. The result of this fusion is a consensus among sources, as opposed to the compromise achieved by many other approaches. Second, our method fully propagates self-localization uncertainty to object-position estimates. We evaluate our method using systematic experiments, which describe an input-error landscape for the performance of our approach. This landscape characterizes how well our method performs when faced with various types and amounts of input errors. ]]> 39 5 1259 1276 1356 Hybrid systems serve as a powerful modeling paradigm for representing complex continuous controlled systems that exhibit discrete switches in their dynamics. The system and the models of the system are nondeterministic due to operation in uncertain environment. Bayesian belief update approaches to stochastic hybrid system state estimation face a blow up in the number of state estimates. Therefore, most popular techniques try to maintain an approximation of the true belief state by either sampling or maintaining a limited number of trajectories. These limitations can be avoided by using bounded intervals to represent the state uncertainty. This alternative leads to splitting the continuous state space into a finite set of possibly overlapping geometrical regions that together with the system modes form configurations of the hybrid system. As a consequence, the true system state can be captured by a finite number of hybrid configurations. A set of dedicated algorithms that can efficiently compute these configurations is detailed. Results are presented on two systems of the hybrid system literature. ]]> 39 5 1277 1291 707 Two strategies for selecting the kernel parameter $(sigma)$ and the penalty coefficient $(C)$ of Gaussian support vector machines (SVMs) are suggested in this paper. Based on viewing the model parameter selection problem as a recognition problem in visual systems, a direct parameter setting formula for the kernel parameter is derived through finding a visual scale at which the global and local structures of the given data set can be preserved in the feature space, and the difference between the two structures can be maximized. In addition, we propose a heuristic algorithm for the selection of the penalty coefficient through identifying the classification extent of a training datum in the implementation process of the sequential minimal optimization (SMO) procedure, which is a well-developed and commonly used algorithm in SVM training. We then evaluate the suggested strategies with a series of experiments on 13 benchmark problems and three real-world data sets, as compared with the traditional 5-cross validation (5-CV) method and the recently developed radius-margin bound (RM) method. The evaluation shows that in terms of efficiency and generalization capabilities, the new strategies outperform the current methods, and the performance is uniform and stable. ]]> 39 5 1292 1307 1135 This paper addresses the ${cal L}_{2}$– ${cal L}_{infty}$ dynamic output feedback (DOF) control problem for a class of nonlinear fuzzy ItÔ stochastic systems with time-varying delay. The focus is placed upon the design of a fuzzy DOF controller guaranteeing a prescribed noise attenuation level in an ${cal L}_{2}$–${cal L}_{infty}$ sense. By using the slack matrix approach, a delay-dependent sufficient condition is derived to assure the mean-square asymptotic stability with an ${cal L}_{2}$ –${cal L}_{infty}$ performance for the closed-loop system. The corresponding solvability condition for a desired ${cal L}_{2}$–${cal L}_{infty}$ DOF controller is established. Since these obtained conditions are not all expressed in terms of linear matrix inequality (LMI), the cone complementary linearization method is exploited to cast them into sequential minimization problems subject to LMI constraints, which can be easily solved numerically. Finally, numerical results are presented to demonstrate the usefulness of the proposed theory. ]]> 39 5 1308 1315 224 A decentralized adaptive methodology is presented for large-scale nonlinear systems with model uncertainties and time-delayed interconnections unmatched in control inputs. The interaction terms with unknown time-varying delays are bounded by unknown nonlinear bounding functions related to all states and are compensated by choosing appropriate Lyapunov–Krasovskii functionals and using the function approximation technique based on neural networks. The proposed memoryless local controller for each subsystem can simply be designed by extending the dynamic surface design technique to nonlinear systems with time-varying delayed interconnections. In addition, we prove that all the signals in the closed-loop system are semiglobally uniformly bounded, and the control errors converge to an adjustable neighborhood of the origin. Finally, an example is provided to illustrate the effectiveness of the proposed control system. ]]> 39 5 1316 1323 259 The multimodal and hierarchical structure characteristics of a system make process modeling quite difficult. In this paper, we present a hierarchical control model (HCM) for hierarchically multimodal processing. From multiple streams, a control layer extracts the inherent group process that denotes the evolution of the system and controls the evolution of every modality. HCMs model the influences of the group on modalities and represent the hierarchical structure of the system by a multilayer network. To estimate the state order of the model, we also present a new information criterion that corrects the preference of traditional criteria for more complex models and proves the rationality of HCMs. Comparisons with other models on multiagent activity recognition show that HCMs are reliable and efficient. ]]> 39 5 1324 1329 685