The cultural algorithm is a means to explicitly acquire problem-solving knowledge from an evolving population and in return apply that knowledge to guide a search.In this paper,the routing problem of mobile agents is formally demonstrated;a model for solving a multi-constrained optimal route is also presented.The cultural algorithm is based on a simulated annealing algorithm and is designed to solve the problem of routing mobile agents.The best individuals,based on Metropolis criterion,are accepted to improve the evolution of the belief space.Experiments showed that the algorithm produces highly competitive results at a relatively low computational cost.

In order to endow dynamic description logic(DDL) with the ability to deal with simple contradiction problems,the extension set and extension transformation were added from extenics,a new discipline which solves contradiction problems.This makes a new kind of description logic,or DDLESET available.The syntax and semantics of DDLESET were developed,then ABoxes in traditional description logic,fuzzy DDL and DDLESET were compared,and finally,the basic reasoning problems in DDLESET were studied and the consistency checking algorithm,or tableau algorithm of traditional description logic was extended.DDLESET possesses the ability in fuzzy expressions and reasoning to assert a positive or negative region dependent degree and provides a solving strategy for assertions whose dependent degree is a negative region.Concepts that could not be satisfied or the actions that could not be realized thus become satisfiable or realizable.

The advantages and characteristics of extension set were presented by comparing extension set with classical set and fuzzy set,and the extension set was introduced as the set theory foundation of dynamic description logic DDL repalcing classical set and fuzzy set to extend DDL,and a new kind of description logic DDLES was proposed,the description form and the semantic explanation of concept,role,instance were given.At last the comparison between the form of Abox in ALC,Fuzzy DDL and DDLES was discussed.

An attribute bagging support vector machine integration method based on independent component analysis (ICA) was developed to improve the generalization performance of support vector machine (SVM). The redundant feature was deleted, and feature subspace was extracted from feature space of independent element, which avoid the destruction of attribute dependence or attribute relativity caused by selecting sub-feature space from original feature space randomly. The performance of single SVM is improved and the diversity between each other is also ensured. Simulations on UCI and StatLog datasets show that the proposed method has better generalization performance.

A hierarchical binary tree multi-class support vector machine (BTMSVM) based on class similarity in feature space is improved to overcome the drawbacks such as unclassifiable region which the existent methods have. The class similarity which considers class distance and distribution sphere in feature space is used to determine the classification order of hierarchical multi-class SVM. The learning samples and corresponding SVM sub-classifier are selectively re-constructed to make sure as bigger as classification margin, as much as generalization ability. The results of simulated experiments show that the proposed method is faster in training and classifying, better in classification correctness and generalization.

Traveling agent problem solves the problem of planning out an optimal migration path when agents migrate to several hosts, which is a complex combinatorial optimization problem. In this paper, an improved ant colony algorithm is presented. A mutation operator is introduced and the local and global updating rules of pheromone are modified on the basis of ant colony algorithm. The algorithm greatly decreases the possibility of falling into stagnation due to arriving at local minimum. The results show that mobile agent can accomplish the computing task with higher efficiency and shorter time.

The base classifiers trained by AdaBoost combination learning algorithm are produced orderly,diversity and complementarity of base classifiers are assured. But along with the iterative process of AdaBoost,the classifier which represents different area classification performance mainly focuses on a certain small area of input space. The constant weights are obtained according to overall classification performance which can not demonstrate base classifiers’ classification performance in different local areas. Based on AdaBoost,a dynamic multiple classifiers combination algorithm based on full information correlation (FIC) which describes base classifiers’ local classification performance is proposed,the classifiers’ selection and their weights are determined according to test samples’ FIC to base classifiers. The simulated experiments show that the combination classification performance is improved greatly.

Traveling agent problem is a complex and combinatorial optimization problem,which solves the problem of finding an optimal path when an agent migrates to several hosts.An improved ant colony algorithm is presented.A mutation operator is introduced.The local and global updating rules of pheromone are modified on the basis of ant colony algorithm with which the possibility of halting the ant system becomes much lower than the ever in the time arriving at local minimum.Experiment shows that the mobile agents can accomplish the computing tasks with much higher efficiency and in a shorter time.

To making description logic can not only check the satisfiability of concept automatically, but also transform the unsatisfied concept to be satisfied, that is to say, endowing the description logic with the ability to dealing with the simple contradiction problem. The theory and method of extenics to solve contradiction problem is referred. The extension set is introduced as the set theory foundation of dynamic description logic DDL replacing classical set and fuzzy set to extend DDL, and a new kind of description logic DDLES is proposed. The description form and the semantic explanation of concept, role, and instance are given. At last, the consistency checking algorithm of Abox is given.

To improve the anti-jamming capability and noise sensitivity of existing support vector machines,and reduce the influence of the outliers on the hyperplane,an improved support vector machine algorithm is proposed based on a weighted adjustable separating hyperplane.In this algorithm,significant values of various training samples were used as weights to assign them to boundary values.Simulation results using the standard UCI data set and an artificial data set show that the proposed algorithm has better ability to resist disturbance and noise and has much more classification precision and requires fewer support vectors compared with a standard or fuzzy support vector machine.

Mobile agent migrates to several hosts for completing its task. Migration strategy is responsible for planning out an optimal migration path, which ensures mobile agent to accomplish its task correctly and efficiently. In this paper, the concept of migration strategy based on support vector machine is proposed. Mobile agent with support vector machine can perceive the changes of environment, react immediately, embodying the reactivity and autonomy of agent. Compared with other migration strategies, it can obtain the optimal result with high probability. The simulation experiment on Aglet platform shows that the migration strategy is effective and available.

Aiming at the problem that Ad-hoc networks mainly choose the shortcut,from the view of complex adaption system,this paper adds the probabilistic reasoning algorithm to Ad-hoc routing protocol,using probabilistic reasoning for nodes’ adaption to choose the route.It solves the congestion and load distribution problems of Ad-hoc networks,improves performance greatly.Using object oriented modeling technique to emulate this new algorithm based on OMNET++ platform,results show that it reduces delay time,enhances QoS,provides a new mode for the development of Ad-hoc networks routing protocol.

The base classifier, which is trained by AdaBoost ensemble learning algorithm, has a constant weight for all test instances. From the view of iterative process of AdaBoost, every base classifier has good classification performance in a certain small area of input space, so the constant weight for different test samples is unreasonable. An improved AdaBoost algorithm based on adaptive weight adjusting is presented. The classifiers' selection and their weights are determined by full information behavior correlation which describes the correlation between test sample and base classifier. The method makes use of all scalars of base classifier's full information behavior, overcomes the problem of information losing. The results of simulated experiments show that the ensemble classification performance is improved greatly.

In this paper,a network intrusion detection method based on PSVMs is constructed.Multiple PSVMs can run at distributed computer system environment.Using feedback to update the initial classifiers,avoid the problem that the learning performance is subject to the distribution state of the data samples in different subsets.Comparison of detection ability between the above detection method and BP neural network,the experiments show that this method can achieve high detection efficiency and low false positive efficiency.And it is superior to traditional BP network method in training time,and has better general ability.

Migration strategy is responsible for planning out an optimal migration path,which ensures that mobile agent can complete its task correctly and efficiently.After analyzing current models and criterions,this paper presents migration strategies based on support vector machine and designs the model.Simulation experiments show that,compared with other migration strategies,it can get the optimal result with high probability,thus to verify this algorithm is effective and available.

Classification has an important role in data mining, but the individual classifier has its limited applicable field, so combining the classified output of multiple classifiers to get much more accuracy is very valuable. There are many combination algorithms such as product, sum, median and vote rules. But these integration algorithms always have not good capability in different datasets. So in this paper a new parallel multiple classifiers combining algorithm, that is Maximum of posterior probability Average with Self-adaptive Weight based on Output vectors and Decision template (MASWOD) is proposed. The experiment on standard UCI dataset show that this algorithm improve the classified accuracy and extend the applicable area of data mining greatly.