Licheng Jiao

A novel genetic algorithim based on immunity

A novel optimal algorithm, immune genetic algorithm (IGA), is proposed based on the theory of immunity in biology, which constructs an immune operator accomplished by two steps, a vaccination and an immune selection. The detail processes of realizing IGA are presented. The methods of selecting vaccines and constructing an immune operator are also given. IGA is illustrated to be able to restrain the degenerate phenomenon evidently during the evolutionary process with examples of TSP, improve the searching capability and efficiency, therefore increase the convergent speed greatly.

The immune quantum-inspired evolutionary algorithm

By leading immune concepts and methods into quantum-inspired evolutionary algorithm (QEA), a novel algorithm, the immune quantum-inspired evolutionary algorithm (IQEA), is proposed. On condition of preserving QEAs advantages, IQEA utilizes some characteristics and knowledge in the pending problems for restraining the repeat and ineffective operations during evolution, so as to improve the algorithm efficiency. The experimental results of the knapsack problem show that the performance of IQEA is superior to the conventional EA (CEA), the immune EA (IEA) and QEA.

A new method for edge detection

A hybrid genetic quantum algorithm (GQA) is proposed for edge detection. GQA is based on the concept and principles of quantum computing such as qubits and superposition of states. By adopting qubit chromosome, GQA can represent a linear superposition of solutions due to its probabilistic representation. Thus, GQA has a better characteristic of diversity and better global search capability than classical approaches. We combine GQA and the local search technique to the problem of edge detection. Experiment results show that the algorithm perform well in terms of quality of the final edge image, rate of convergence and robustness to noise.

An edge detector based on parallel quantum-inspired evolutionary algorithm

This work proposes a hybrid parallel quantum-inspired evolutionary algorithm (PQEA) based on cost minimization technique for edge detection. Quantum-inspired evolutionary algorithm (QEA) is based on the concepts and principles of quantum computing such as qubits and superposition of states. By adopting qubit chromosome as a representation, QEA can represent a linear superposition of solutions due to its probabilistic representation. QEA is more suitable for parallel structure than the conventional evolutionary algorithms because of rapid convergence and good global search capability. We combine PQEA and the local search technique to solve the problem of edge detection. Experimental results show that the algorithm perform very well in terms of the quality of the final edge image, rate of convergence and robustness to noise.

FWNN for interval estimation with interval learning algorithm

In this paper, a wavelet-based fuzzy neural network for interval estimation of processed data with its interval learning algorithm is proposed. It is also proved to be an efficient approach to calculate the wavelet coefficient.

Ship target recognition using kernel Fisher discriminant

The classification of ship targets using the kernel Fisher discriminant analysis is investigated in this paper. The main idea of this method is to find a nonlinear direction by first mapping the data nonlinearly into some feature space and compute Fishers linear discriminant in input space. Based on the kernel Fisher discriminant, we recognize three types of ships. The satisfactory experimental results are obtained. In addition, we compare this method with other state of the art classification techniques. The experiments show that the kernel Fisher discriminant is superior to the other algorithms.

An adaptive neurofuzzy network for identification of the complicated nonlinear system

This paper presents a compound neural network model, i.e., adaptive neurofuzzy network (ANFN), which can be used for identifying the complicated nonlinear system. The proposed ANFN has a simple structure and exploits a hybrid algorithm combining supervised learning and unsupervised learning. In addition, ANFN is capable of overcoming the error of system identification due to the existence of some changing points and improving the accuracy of identification of the whole system. The effectiveness of the model and its algorithm is tested on the identification results of missile attacking area.

Fuzzy Wavelet Neural Networks: Theory and Applications

The fuzzy wavelet neural networks (FWNN) are proposed in this paper. The structure and two learning algorithms of the FWNN for R-F function are given. Under the framework of such structure and two learning algorithms, wavelet-based fuzzy neural networks for interval estimation of processed data and for interpolation of fuzzy if-then rules are proposed too. The simulation results are given to prove their feasibility.

Neural networks via nonlinear sensitivity approach

The sensitivity, semisensitivity, trail sensitivity, and nth-order zero sensitivity of neural networks are investigated. The sensitivity analysis approach of nonlinear neural network systems is established. As an example, the sensitivities of the Hopfield neural network model are studied.<<ETX>>

An efficient method of automatical feature extraction and target classification

P.R. Chinas fishery and offshore petroleum development have been in urgent need of a classifier of noise signals. In this paper, the feature extraction mechanism and the classification mechanism of the adaptive wavelet neural network are revealed by mathematical analysis, and an efficient engineering classifier based on the adaptive wavelet neural network is designed and applied to classify the actual ship noises. The classification experiment results show that the mathematical analysis of the feature extraction mechanism and the classification mechanism is correct and provides theoretical bases for the adaptive wavelet neural network classifier design.