Jiao Licheng

Optimal approximation of linear systems by artificial immune response

This paper puts forward a novel artificial immune response algorithm for optimal approximation of linear systems. A quaternion model of artificial immune response is proposed for engineering computing. The model abstracts four elements, namely, antigen, antibody, reaction rules among antibodies, and driving algorithm describing how the rules are applied to antibodies, to simulate the process of immune response. Some reaction rules including clonal selection rules, immunological memory rules and immune regulation rules are introduced. Using the theorem of Markov chain, it is proofed that the new model is convergent. The experimental study on the optimal approximation of a stable linear system and an unstable one show that the approximate models searched by the new model have better performance indices than those obtained by some existing algorithms including the differential evolution algorithm and the multi-agent genetic algorithm.

Automatic parameters selection for SVM based on GA

Motivated by the fact that automatic parameter selection for support vector machines (SVM) is an important issue in order to make the SVM practically useful against the commonly used leave-one-out (loo) method, which has complex calculation and time consuming. An effective strategy for automatic parameter selection for SVM is proposed by using the genetic algorithm (GA) in this paper. Simulation results of the practice data model demonstrate the effectiveness and high efficiency of the proposed approach.

Intelligent particle swarm optimization in multiobjective optimization

How to find a sufficient number of uniformly distributed and representative Pareto optimal solutions is very important for Multiobjective Optimization (MO) problems. A new model for Particle Swarm Optimization is constructed firstly, and then an Intelligent Particle Swarm Optimization (IPSO) for MO problems is proposed based on AER (Agent-Environment-Rules) model, in which competition operator and clonal slection operator are designed to provide an appropriate selection pressure to propel the swarm population towards the Pareto-optimal front. The quantitative and qualitative comparisons indicate that the proposed approach is highly competitive and that can be considered as a viable alternative to solve MO problems.How to find a sufficient number of uniformly distributed and representative Pareto optimal solutions is very important for multiobjective optimization (MO) problems. A new model for particle swarm optimization is constructed firstly, and then an intelligent particle swarm optimization (IPSO) for MO problems is proposed based on AER (agent-environment-rules) model, in which competition operator and clonal selection operator are designed to provide an appropriate selection pressure to propel the swarm population towards the Pareto-optimal front. The quantitative and qualitative comparisons indicate that the proposed approach is highly competitive and that can be considered as a viable alternative to solve MO problems

A noval algorithm of artificial immune system for high-dimensional function numerical optimization∗

Abstract Based on the clonal selection theory and immune memory theory, a novel artificial immune system algorithm, immune memory clonal programming algorithm (IMCPA), is put forward. Using the theorem of Markov chain, it is proved that IMPCA is convergent. Compared with some other evolutionary programming algorithms (like Breeder genetic algorithm), IMPCA is shown to be an evolutionary strategy capable of solving complex machine learning tasks, like high-dimensional function optimization, which maintains the diversity of the population and avoids prematurity to some extent, and has a higher convergence speed. ∗Supported by National Natural Science Foundation of China (Grant Nos. 60133010 and 60372045)

A GA-based clustering algorithm for large data sets with mixed and categorical values

In the field of data mining, it is often encountered to perform cluster analysis on large data sets with mixed numeric and categorical values. However, most existing clustering algorithms are only efficient for the numeric data rather than the mixed data set. For this purpose, this paper presents a novel clustering algorithm for these mixed data sets by modifying the common cost function, trace of the within cluster dispersion matrix. The genetic algorithm (GA) is used to optimize the new cost function to obtain valid clustering result. Experimental result illustrates that the GA-based new clustering algorithm is feasible for the large data sets with mixed numeric and categorical values.

Immunity clonal strategies

Based on the clonal selection theory, the main mechanisms of clone, which will be explored in the field of artificial intelligence, are analyzed in this paper. An improved evolutionary strategy algorithm, immunity clonal strategy algorithm (ICS), which includes immunity monoclonal strategy algorithm (IMSA) and immunity polyclonal strategy algorithm (IPSA), is put forward. Compared with the classical evolutionary strategy algorithm (CES), ICS is shown to be an evolutionary strategy capable of solving complex machine learning tasks, like multi-objective optimization, and the results are better. Using the theories of Markov chain, it is proved that ICS algorithm is convergent.

Adaptive chaos clonal evolutionary programming algorithm

Based on the chaos movement and the clonal selection theory, a novel artificial immune system algorithm, Adaptive Chaos Clonal Evolutionary Programming Algorithm (ACCEP), is proposed in this paper. The new algorithm uses the Logistic Sequence to control the mutation scale and uses the Chaos Mutation Operator to control the clonal selection. Compared with SGA and Clonal Selection Algorithm, ACCEP can enhance the precision and stability, avoid prematurity to some extent, and have the high convergence speed. The results of the experiment indicate that ACCEP has the capability to solve complex machine learning tasks, like Multimodal Function Optimization.

Radar target recognition based on support vector machine

Support vector machine is a new kind technique for pattern recognition. We study the classification mechanism of SVM in detail in this paper. Finally, the simulations are done according to SVM algorithm and the results show the advantage of it.

Classification mechanism of support vector machines

The purpose of this paper is to provide an introductory tutorial on the basic ideas behind support vector machines (SVM). The paper starts with an overview of structural risk minimization (SRM) principle, and describes the mechanism of how to construct SVM. For a two-class pattern recognition problem, we discuss in detail the classification mechanism of SVM in three cases of linearly separable, linearly nonseparable and nonlinear. Finally, for nonlinear case, we give a new function mapping technique: By choosing an appropriate kernel function, the SVM can map the low-dimensional input space into the high dimensional feature space, and construct an optimal separating hyperplane with maximum margin in the feature space.

Adaptive Synchronization between Two Different Complex Networks with Time-Varying Delay Coupling

A new general network model for two complex networks with time-varying delay coupling is presented. Then we investigate its synchronization phenomena. The two complex networks of the model differ in dynamic nodes, the number of nodes and the coupling connections. By using adaptive controllers, a synchronization criterion is derived. Numerical examples are given to demonstrate the effectiveness of the obtained synchronization criterion. This study may widen the application range of synchronization, such as in chaotic secure communication.