editorial artificial intelligence and its applications...

EditorialArtificial Intelligence and Its Applications 2014

Yudong Zhang,1 Saeed Balochian,2 Praveen Agarwal,3 Vishal Bhatnagar,4

and Orwa Jaber Housheya5

1School of Computer Science and Technology, Nanjing Normal University, Nanjing, Jiangsu 210023, China2Department of Electrical Engineering, Islamic Azad University, Gonabad Branch, Gonabad 96916-29, Iran3Department of Mathematics, Anand International College of Engineering, Jaipur 303012, India4Ambedkar Institute of Advanced Communication Technologies and Research, Government of NCT of Delhi,Geeta Colony, Delhi 110031, India5Department of Chemistry, Arab American University, Jenin, State of Palestine

Correspondence should be addressed to Yudong Zhang; [email protected]

Received 22 February 2016; Accepted 23 February 2016

Copyright © 2016 Yudong Zhang et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Artificial intelligence (AI) researches the intelligence exhib-ited by machines. It creates revolutionized information tech-nology. The world famous companies like Google, Yahoo,Facebook, Baidu, and so forth have spent millions of dollarsto research on developing new algorithms on AI. Neverthe-less, there are a number of challenging issues in realisticapplications due to fast-growing large and complex problems.

This special issue aims to bring together academia andindustry experts to report on the recent developments onartificial intelligence and its applications, in every aspect ofartificial intelligence technology, includingmachine learning,data mining, computer vision, multiagent systems, evolu-tionary computation, deep learning, and fuzzy logic. Theprimary guideline was to either demonstrate the most signif-icant developments on the topics of AI or apply AI-relatedalgorithms in real-life scenarios.

In the paper entitled “A Comprehensive Survey onParticle Swarm Optimization Algorithm and Its Applica-tions,” Y. Zhang et al. investigated the particle swarm opti-mization (PSO). Their survey presented a comprehensiveinvestigation of PSO. On one hand, they provided advanceswith PSO, including its modifications (including quantum-behaved PSO, bare-bones PSO, chaotic PSO, and fuzzy PSO),population topology (as fully connected, vonNeumann, ring,star, random, etc.), hybridization (with genetic algorithm,simulated annealing, Tabu search, artificial immune sys-tem, ant colony algorithm, artificial bee colony, differential

evolution, harmonic search, and biogeography-based opti-mization), extensions (to multiobjective, constrained, dis-crete, and binary optimization), theoretical analysis (param-eter selection and tuning and convergence analysis), andparallel implementation (in multicore, multiprocessor, GPU,and cloud computing forms). On the other hand, theyoffered a survey on applications of PSO to the followingeight fields: electrical and electronic engineering, automationcontrol systems, communication theory, operations research,mechanical engineering, fuel and energy, medicine, chem-istry, and biology. It is hoped that this survey would bebeneficial for the researchers studying PSO algorithms.

In the paper entitled “Model of Multilayer KnowledgeDiffusion for Competence Development in an Organiza-tion,” P. Rozewski and J. Jankowski proposed several modelswith the main goal of simulating diffusion and explainingthe nature of knowledge diffusion. They extended existingapproaches by using multilayer diffusion model and focusedon analysis of competence development process. The pro-posed model described competence development processin a new way through horizontal and vertical knowledgediffusion in multilayer network. In the network, agentscollaborated and interchanged various kinds of knowledgethrough different layers and these mutual activities affectedthe competencies in a positive or negative way. Taking intoconsideration worker’s cognitive and social abilities and theprevious level of competence, the new competence level

Hindawi Publishing CorporationMathematical Problems in EngineeringVolume 2016, Article ID 3871575, 6 pageshttp://dx.doi.org/10.1155/2016/3871575

2 Mathematical Problems in Engineering

can be estimated. Their model was developed to supportcompetence management in different organizations.

In the paper entitled “Landslide Occurrence PredictionUsing Trainable Cascade Forward Network and MultilayerPerceptron,” M. S. Al-batah et al. introduced two modelsof artificial neural network, namely, multilayer perceptron(MLP) and cascade forward neural network (CFNN), topredict the landslide hazard map of Penang Island. Thesetwomodels were tested and compared using elevenmachine-learning algorithms: LevenbergMarquardt, Broyden FletcherGoldfarb, resilient backpropagation, scaled conjugate gradi-ent, conjugate gradient with Beale, conjugate gradient withFletcher Reeves updates, conjugate gradient with Polakribiereupdates, one step secant, gradient descent, gradient descentwith momentum and adaptive learning rate, and gradientdescent with momentum algorithm. The performance of thelandslide prediction depends on the input factors besidethe prediction method. In this research work, 14 inputfactors were used.The prediction accuracies of networkswereverified using the area under curve for the receiver operatingcharacteristics. The results indicated that the best predictionaccuracy was achieved using the CFNN network with theLevenberg Marquardt learning algorithm, that is, 82.89% forthe training data set and 81.62% for the testing data set.

In the paper entitled “Optimization of Train Trip Pack-age Operation Scheme,” L. Tong et al. firstly analyzed therelated factors of train trip package transportation fromits organizational forms and characteristics. Then an opti-mization model for train trip package transportation wasestablished to provide optimum operation schemes. Theproposed model was solved by the genetic algorithm. Atlast, the paper tested the model on the basis of the dataof 8 regions. The results showed that the proposed methodis feasible for solving operation scheme issues of train trippackage.

In the paper entitled “A Self-Adaptive Hidden MarkovModel for Emotion Classification in Chinese Microblogs,”L. Liu et al. proposed a modified version of hidden Markovmodel (HMM) classifier, called self-adaptive HMM, theparameters of which were optimized by particle swarmoptimization algorithms. Since manually labeling large-scaledataset was difficult, Liu et al. also employed the entropy todecide whether a new unlabeled tweet will be contained inthe training dataset after being assigned an emotion using theproposed HMM-based approach. In the experiment, about200,000 Chinese tweets from SinaWeibo were collected. Theresults showed that the 𝐹-score of the approach gets 76% onhappiness and fear and 65% on anger, surprise, and sadness.In addition, the self-adaptive HMM classifier outperformedNaive Bayes and support vector machine on recognition ofhappiness, anger, and sadness.

In the paper entitled “Automatic Classification of RemoteSensing Images Using Multiple Classifier Systems,” B. Yanget al. aimed to correctly identify land use types reflected inremote sensing images. Support vector machine, maximumlikelihood classifier, backpropagation neural network, fuzzy𝑐-means, and minimum distance classifier were combinedto construct three multiple classifier systems (MCSs). TwoMCSs were implemented, namely, comparative major voting

(CMV) and Bayesian average (BA). One method called WA-AHP was proposed, which introduced analytic hierarchyprocess intoMCS. Classification results of base classifiers andMCSs were compared with the ground-truth map. Accuracyindicators were computed and receiver operating character-istic curves were illustrated, so as to evaluate the performanceof MCSs. Experimental results showed that employing MCSscan increase classification accuracy significantly, comparedwith base classifiers. From the accuracy evaluation resultand visual check, the best MCS is WA-AHP with overallaccuracy of 94.2%, which overmatches BA and rivals CMV.Theproducer’s accuracy of each land use type proves the goodperformance ofWA-AHP.Therefore, MCS is superior to baseclassifiers in remote sensing image classification, and WA-AHP is an efficient MCS.

In the paper entitled “A Novel Tournament SelectionBased Differential Evolution Variant for Continuous Opti-mization Problems,” Q. Abbas et al. proposed a noveltournament based parent selection variant of differentialevolution (DE) algorithm. The proposed variant enhancedsearching capability and improved convergence speed ofDE algorithm. This paper also presented a novel statisticalcomparison of existing DE mutation variants, which catego-rized these variants in terms of their overall performance.Experimental results showed that the proposed DE vari-ant had significance performance over other DE mutationvariants.

In the paper entitled “Manifold Learning with Self-Organizing Mapping for Feature Extraction of NonlinearFaults in Rotating Machinery,” L. Liang et al. proposeda new method for extracting the low-dimensional featureautomatically with self-organization mapping manifold forthe detection of rotating mechanical nonlinear faults (suchas rubbing and pedestal looseness). Under the phase spacereconstructed by single vibration signal, the self-organizationmapping (SOM) with expectation maximization iterationalgorithm was used to divide the local neighborhoods adap-tively without manual intervention. After that, the localtangent-space alignment algorithm was adopted to compressthe high-dimensional phase space into low-dimensionalfeature space. The proposed method takes advantages ofthe manifold learning in low-dimensional feature extractionand adaptive neighborhood construction of SOM and canextract intrinsic fault features of interest in two-dimensionalprojection space. To evaluate the performance of the pro-posed method, the Lorenz system was simulated and rota-tion machinery with nonlinear faults was obtained for testpurposes. Compared with the holospectrum approaches, theresults reveal that the proposed method is superior in iden-tifying faults and effective for rotating machinery conditionmonitoring.

In the paper entitled “Modifying Regeneration Muta-tion and Hybridising Clonal Selection for EvolutionaryAlgorithms Based Timetabling Tool,” T. Thepphakorn etal. outlined the development of a new evolutionary algo-rithms based timetabling (EAT) tool for solving coursescheduling problems that include a genetic algorithm (GA)and a memetic algorithm (MA). Reproduction processesmay generate infeasible solutions. Previous research used

Mathematical Problems in Engineering 3

repair processes that were applied after a population ofchromosomes was generated. This research developed a newapproach which (i) modified the genetic operators to preventthe creation of infeasible solutions before chromosomes wereadded to the population and (ii) included the clonal selectionalgorithm (CSA) and the elitist strategy (ES) to improvethe quality of the solutions produced. This approach wasadopted by both the GA and MA within the EAT. TheMA was further modified to include hill climbing localsearch. The EAT program was tested using 14 benchmarktimetabling problems from the literature using a sequentialexperimental design, which included a fractional factorialscreening experiment. Experiments were conducted to (i)test the performance of the proposed modified algorithms,(ii) identify which factors and interactions were statisticallysignificant, (iii) identify appropriate parameters for the GAand MA, and (iv) compare the performances of the varioushybrid algorithms. The genetic algorithm with modifiedgenetic operators produced an average improvement of over50%.

In the paper entitled “Fuzzy Wavelet Neural NetworkUsing a Correntropy Criterion for Nonlinear System Iden-tification,” L. L. S. Linhares et al. investigated the fuzzywavelet neural networks (FWNNs) that are an efficient toolto identify nonlinear systems. In these structures, featuresrelated to fuzzy logic, wavelet functions, and neural networksare combined in an architecture similar to the adaptiveneurofuzzy inference systems (ANFIS). In practical applica-tions, the experimental data set used in the identificationtask often contains unknown noise and outliers, whichdecrease the FWNN model reliability. In order to reduce thenegative effects of these erroneous measurements, their workproposed the direct use of a similarity measure based oninformation theory in the FWNN learning procedure. Themean squared error (MSE) cost function was replaced bythemaximum correntropy criterion (MCC) in the traditionalerror backpropagation (BP) algorithm. The input-outputmaps of a real nonlinear system studied were identified froman experimental data set corrupted by different outliers ratesand additive white Gaussian noise. The results demonstratedthe advantages of the proposed cost function using the MCCas compared to the MSE. This work also investigated theinfluence of the kernel size on the performance of the MCCin the BP algorithm, since it is the only free parameter ofcorrentropy.

In the paper entitled “A Novel WLAN Client Puzzleagainst DoS Attack Based on Pattern Matching,” A. Ordiet al. analyzed the popularity of 802.11 based networks andpointed out that they suffered several types of DoS attack,launched by an attacker whose aim is to make an accesspoint (AP) unavailable to legitimate users. One of the mostcommon DoS attacks on 802.11 based networks is to depletethe resources of the AP. A serious situation like this canoccur when the AP receives a burst of connection requests.Their paper addressed this commonDoS attack and proposeda lightweight puzzle, based on pattern matching. Using apattern-matching technique, this model adequately resistedresource-depletion attacks in terms of both puzzle generationand solution verification. Using a sensible series of contextual

comparisons, the outcomes were modelled by a simulator,and the security definition and proofs are verified, amongother results.

In the paper entitled “Weight Optimization in RecurrentNeural Networks with Hybrid Metaheuristic Cuckoo SearchTechniques for Data Classification,” N.M. Nawi et al. investi-gated recurrent neural network (RNN). This network can beeducated with gradient descent backpropagation. However,traditional training algorithms had some drawbacks such asslow speed of convergence being not definite to find the globalminimum of the error function since gradient descent mayget stuck in localminima.As a solution, nature inspiredmeta-heuristic algorithms provide derivative-free solution to opti-mize complex problems. This paper proposed a new meta-heuristic search algorithm called Cuckoo Search (CS) basedon Cuckoo bird’s behavior to train Elman recurrent net-work (ERN) and backpropagation Elman recurrent network(BPERN) in achieving fast convergence rate and to avoidlocal minima problem.The proposed CSERN and CSBPERNalgorithms were compared with artificial bee colony usingBP algorithm and other hybrid variants algorithms. Specifi-cally, some selected benchmark classification problems wereused. The simulation results showed that the computa-tional efficiency of ERN and BPERN training process washighly enhanced when coupled with the proposed hybridmethod.

In the paper entitled “StereoMatching Based on ImmuneNeural Network in Abdomen Reconstruction,” H. Liu etal. suggested stereo feature matching is a technique thatfinds an optimal match in two images from the same entityin the three-dimensional world. The stereo correspondenceproblem is formulated as an optimization task where anenergy function, which represents the constraints on thesolution, is to be minimized. They proposed a novel intelli-gent biological network (Bio-Net), which involves the humanB-T cells immune system into neural network, in order tolearn the robust relationship between the input feature pointsand the output matched points. In the experiments, theabdomen reconstructions for different-shape mannequinswere performed by means of the proposed method. The finalresults were compared and analyzed, which demonstrate thatthe proposed approach greatly outperforms the single neuralnetwork and the conventional matching algorithm in precise.Particularly, with respect to time cost and efficiency, theproposed method exhibits its significant promising potentialfor improvement. Hence, it is entirely considered as aneffective and feasible alternative option for stereo matching.

In the paper entitled “Neural Network-Based Fault-Tolerant Control of Underactuated Surface Vessels,” B.S. Park addressed the problem of trajectory trackingof underactuated surface vessels (USVs) in the presenceof thruster failure. Multilayer neural networks (MNNs)were employed to estimate the unknown model parame-ters and external disturbances. To design a fault-tolerantcontroller without a fault detection scheme, they usedthe Nussbaum gain technique. They also introduced anadditional control to resolve the difficulty arising fromhaving fewer inputs than degrees of freedom. Further,an approach angle was proposed to track both straight


and curved path. Stability analysis and simulations wereperformed to demonstrate the effectiveness of the proposedscheme.

In the paper entitled “Dynamic Track Management inMHT for Pedestrian Tracking Using Laser Range Finder,” A.H. A. Rahman et al. proposed a multilevel clustering of laserrange finder (LRF) data to improve the accuracy of a trackingsystem by adding another clustering level after the featureextraction process. A dynamic track management (DTM)was introduced in multiple hypothesis tracking (MHT) withmultiple motion models to perform a track creation, associa-tion, and deletion. The experimental results from real timeimplementation proved that the proposed multiclusteringis capable of producing a better performance with lesscomputational complexity for a track management process.The proposed dynamic track management is able to solve thetracking problemwith lower computation time when dealingwith occlusion, crossed track, and track deletion.

In the paper entitled “A Study on Many-Objective Op-timization Using the Kriging-Surrogate-Based EvolutionaryAlgorithm Maximizing Expected Hypervolume Improve-ment,” C. Luo et al. investigated and compared the many-objective optimization performance of the Kriging-sur-rogate-based evolutionary algorithm (EA), which maximizesexpected hypervolume improvement (EHVI) for updatingthe Kriging model, with those using expected improvement(EI) and estimation (EST) updating criteria. Numericalexperiments were conducted in 3- to 15-objective DTLZ1-7 problems. An exact hypervolume calculating algorithmwas used for the problems with less than six objectives. Onthe other hand, an approximate hypervolume calculatingalgorithm based on Monte Carlo sampling was adopted forthe problems with more objectives. The results indicate that,in the nonconstrained case, EHVI is a highly competitiveupdating criterion for theKrigingmodel andEAbasedmany-objective optimization, especially when the test problem iscomplex and the number of objectives or design variables islarge.

In the paper entitled “System Optimization for Tempo-ral Correlated Cognitive Radar with EBPSK-Based MCPCSignal,” P. Chen and L. Wu proposed a novel radar workingscheme to consider both target detection and estimation.At the detection stage, the generalized likelihood ratio test(GLRT) threshold was deduced, and the GLRT detectionprobability was given. At the estimation stage, an approachbased on Kalman filtering (KF) was proposed to estimatetarget scattering coefficients (TSC), and the estimation per-formance was improved significantly by exploiting the TSCtemporal correlation. Additionally, the optimal waveformwas obtained to minimize the mean squared error (MSE)of KF estimation. For the practical consideration, itera-tion algorithms were proposed to optimize the EBPSK-based multicarrier phase-coded (MCPC) signal in termsof power allocation and coding matrix. Simulation resultsdemonstrated that the KF estimation approach can improvethe estimation performance by 25% compared with max-imum a posteriori probability (MAP) method, and theKF estimation performance can be further improved by90% by optimizing the transmitted waveform spectrum.

Moreover, by optimizing the power allocation and codingmatrix of the EBPSK-based MCPC signal, the KF estima-tion performances were, respectively, improved by 7% and8%.

In the paper entitled “Color Image EncryptionAlgorithmBased on TD-ERCS System and Wavelet Neural Network,”K. Zhang and J. Fang proposed a new image encryptionalgorithm based on TD-ERCS system andwavelet neural net-work, in order to solve the security problem of transmissionimage across public networks. According to the permutationprocess and the binary XORoperation from the chaotic seriesby producing TD-ERCS system and wavelet neural network,it can achieve image encryption. This encryption algorithmwas a reversible algorithm, and it can achieve original imagein the rule inverse process of encryption algorithm. Finally,through computer simulation, the experiment results showedthat the new chaotic encryption algorithm based on TD-ERCS system and wavelet neural network is valid and hashigher security.

In the paper entitled “Fault Diagnosis of Supervision andHomogenizationDistance Based on Local Linear EmbeddingAlgorithm,” G. Wang et al. developed an improved locallinear embedding algorithm of homogenization distance(HLLE), in view of the problems of uneven distribution ofreality fault samples and dimension reduction effect of locallylinear embedding (LLE) algorithm, which is easily affected byneighboring points. The method made the overall distribu-tion of sample points tend to be homogenization and reducesthe influence of neighboring points using homogenizationdistance instead of the traditional Euclidean distance. It washelpful to choose effective neighboring points to constructweight matrix for dimension reduction. Because the faultrecognition performance improvement of HLLE was limitedand unstable, the paper further proposed a new local linearembedding algorithm of supervision and homogenizationdistance (SHLLE) by adding the supervised learning mech-anism. On the basis of homogenization distance, super-vised learning increased the category information of samplepoints, so that the same category of sample points will begathered and the heterogeneous category of sample pointswill be scattered. It effectively improved the performance offault diagnosis and maintains stability at the same time. Acomparison of the methods mentioned above was made bysimulation experiment with rotor system fault diagnosis, andthe results showed that SHLLE algorithm has superior faultrecognition performance.

In the paper entitled “A New Asymptotic Notation:Weak Theta,” A.-H. Mogos et al. defined a new asymptoticnotation, called “Weak Theta,” that used the comparisonof various complexity functions with two given complexityfunctions. Weak Theta notation was especially useful incharacterizing complexity functions whose behavior washard to be approximated using a single complexity func-tion. In addition, in order to highlight the main particu-larities of Weak Theta, they proposed and proved severaltheoretical results: properties of Weak Theta, criteria forcomparing two complexity functions, and properties of anew set of complexity functions based on Weak Theta.Furthermore, to illustrate the usefulness of this notation,

Mathematical Problems in Engineering 5

the authors discussed an application of Weak Theta inartificial intelligence.

In the paper entitled “Locating High-Impedance FaultSection in Electric Power Systems Using Wavelet Transform,𝑘-Means, Genetic Algorithms, and Support VectorMachine,”Y.-Y. Hong and W.-S. Huang presented a new method forlocating the line (feeder) section of the high-impedance fault(HIF) with the help of limited measurements in electricpower systems. The discrete wavelet transform was used toextract the features of transients caused by HIFs. A modified𝑘-means algorithm associated with genetic algorithms wasthen utilized to determine the placement of measurementfacilities. The signal energies attained by wavelet coefficientsserved as inputs to the support vector machine for locatingthe HIF line section.The simulation results obtained from an18-busbar distribution system showed the applicability of theproposed method.

In the paper entitled “Recognition of Mixture ControlChart Pattern Using Multiclass Support Vector Machine andGenetic Algorithm Based on Statistical and Shape Features,”M. Zhang and W. Cheng introduce an intelligent hybridmodel for recognizing the mixture control chart patterns(CCPs) that included three main aspects: feature extrac-tion, classifier, and parameters optimization. In the featureextraction, statistical and shape features of observation datawere used in the data input to get the effective data for theclassifier. A multiclass support vector machine (MSVM) wasapplied for recognizing the mixture CCPs. Finally, geneticalgorithm (GA) was utilized to optimize theMSVM classifierby searching the best values of the parameters of MSVM andkernel function. The performance of the hybrid approachwas evaluated by simulation experiments, and simulationresults demonstrated that the proposed approach was able toeffectively recognize mixture CCPs.

In the paper entitled “A Fault-Tolerant Filtering Algo-rithm for SINS/DVL/MCP Integrated Navigation System,”X. Xu et al. proposed a fault-tolerant adaptive Kalmanfilter (FTAKF) algorithm for the integrated navigation sys-tem composed of a strapdown inertial navigation system(SINS), a Doppler velocity log (DVL), and a magneticcompass (MCP). The evolutionary artificial neural networks(EANN) were used in self-learning and training of theintelligent data fusion algorithm. The proposed algorithmcan significantly outperform the traditional KF in pro-viding estimation continuously with higher accuracy andsmoothing the KF outputs when observation data wereinaccurate or unavailable for a short period.The experimentsof the prototype verified the effectiveness of the proposedmethod.

In the paper entitled “Multiagent Cooperative LearningStrategies for Pursuit-Evasion Games,” J. Y. Kuo et al. exam-ined the pursuit-evasion problem for coordinating multiplerobotic pursuers to locate and track a nonadversarial mobileevader in a dynamic environment. Two kinds of pursuitstrategies were proposed, one for agents that cooperatewith each other and the other for agents that operateindependently. This work further employed the probabilistictheory to analyze the uncertain state information about thepursuers and the evaders and uses case-based reasoning to

equip agents withmemories and learning abilities. Accordingto the concepts of assimilation and accommodation, bothpositive-angle and bevel-angle strategies were developed toassist agents in adapting to their environment effectively. Thecase study analysis used the recursive porous agent simu-lation toolkit (REPAST) to implement a multiagent systemand demonstrated superior performance of the proposedapproaches to the pursuit-evasion game.

In the paper entitled “User Adapted Motor-ImaginaryBrain-Computer Interface by means of EEG Channel Selec-tion Based on Estimation of Distributed Algorithms,” A.Astigarraga et al. presented a personalized interface designmethod, for electroencephalogram (EEG) based Brain-Computer Interfaces (BCIs), based on channel selection.They described a novel two-step method in which firstlya computationally inexpensive greedy algorithm found anadequate search range; and then, an estimation of distributionalgorithm (EDA) was applied in the reduced range to obtainthe optimal channel subset. The use of the EDA algorithmallowed us to select the most interacting channels subset,removing the irrelevant and noisy ones, thus selecting themost discriminative subset of channels for each user improv-ing accuracy. The method was tested on the IIIa datasetfrom the BCI competition III. Experimental results showedthat the resulting channel subset was consistent with motorimaginary related neurophysiological principles and, on theother hand, optimized performance reducing the number ofchannels.

In the paper entitled “KD-ACP: A Software Frameworkfor Social Computing in Emergency Management,” B. Chenet al. addressed the application of a computational theory andrelated techniques for studying emergency management insocial computing.They proposed a novel software frameworkcalled KD-ACP. The framework provided a systematic andautomatic platform for scientists to study the emergencymanagement problems in three aspects:modelling the societyin emergency scenario as the artificial society; investigatingthe emergency management problems by the repeat com-putational experiments; parallel execution between artificialsociety and the actual society managed by the decisionsfrom computational experiments. The software frameworkwas composed of a series of tools. These tools were cat-egorized into three parts corresponding to “A,” “C,” and“P,” respectively. Using H1N1 epidemic in Beijing city as thecase study, the modelling and data generating of Beijingcity, experiments with settings of H1N1, and interventionmeasures and parallel execution by situation tool wereimplemented by KD-ACP. The results output by the softwareframework showed that the emergency response decisionscan be tested to find a more optimal one through thecomputational experiments. In the end, the advantages ofthe KD-ACP and the future work were summarized in theconclusion.

Acknowledgments

Wewould like to showour appreciation to all authors for theircontributions and the reviewers for their effort providing


valuable and prompt comments. We hope that this specialissue offers a comprehensive and timely view of the advancesin artificial intelligence and its applications, and we expectthat it will offer stimulation for further research.

Yudong ZhangSaeed BalochianPraveen AgarwalVishal Bhatnagar

Orwa Jaber Housheya

Submit your manuscripts athttp://www.hindawi.com

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

MathematicsJournal of


Mathematical Problems in Engineering

Hindawi Publishing Corporationhttp://www.hindawi.com

Differential EquationsInternational Journal of

Volume 2014

Applied MathematicsJournal of


Probability and StatisticsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Journal of


Mathematical PhysicsAdvances in

Complex AnalysisJournal of


OptimizationJournal of


CombinatoricsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

International Journal of


Operations ResearchAdvances in

Journal of


Function Spaces

Abstract and Applied AnalysisHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

International Journal of Mathematics and Mathematical Sciences


The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014


Algebra

Discrete Dynamics in Nature and Society



Decision SciencesAdvances in

Discrete MathematicsJournal of

Hindawi Publishing Corporationhttp://www.hindawi.com

Volume 2014 Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Stochastic AnalysisInternational Journal of

editorial artificial intelligence and its applications...

Documents