When confronting the complex problems, radial basis function (RBF) neural network has the advantages of adaptive and self-learning ability, but it is difficult to determine the number of hidden layer neurons, and the weights learning ability from hidden layer to the output layer is low; these deficiencies easily lead to decreasing learning ability and recognition precision. Aiming at this problem, we propose a new optimized RBF neural network algorithm based on genetic algorithm (GA-RBF algorithm), which uses genetic algorithm to optimize the weights and structure of RBF neural network; it chooses new ways of hybrid encoding and optimizing simultaneously. Using the binary encoding encodes the number of the hidden layer's neurons and using real encoding encodes the connection weights. Hidden layer neurons number and connection weights are optimized simultaneously in the new algorithm. However, the connection weights optimization is not complete; we need to use least mean square (LMS) algorithm for further leaning, and finally get a new algorithm model. Using two UCI standard data sets to test the new algorithm, the results show that the new algorithm improves the operating efficiency in dealing with complex problems and also improves the recognition precision, which proves that the new algorithm is valid.

This paper reviews the use of evolutionary algorithms (EAs) to optimize artificial neural networks (ANNs). First, we briefly introduce the basic principles of artificial neural networks and evolutionary algorithms and, by analyzing the advantages and disadvantages of EAs and ANNs, explain the advantages of using EAs to optimize ANNs. We then provide a brief survey on the basic theories and algorithms for optimizing the weights, optimizing the network architecture and optimizing the learning rules, and discuss recent research from these three aspects. Finally, we speculate on new trends in the development of this area.

In research of pattern recognition, we always want to achieve the correct classification rate according to the characteristics required. Feature extraction greatly affects the design and performance of the classifier, and it is one of the core issue of PR research. As an important component of pattern recognition, feature extraction has been paid close attention by many scholars, and currently has become one of the research hot spots in the field of pattern recognition. This article gives a general discussion of feature extraction, includes linear feature extraction and nonlinear feature extraction, and introduces the frontier methods of this field, at last discusses the development tendency of feature extraction.

In the traditional learning algorithms of radial basis function (RBF) neural network, the architecture of the network is hard to be decided; thereby, the learning ability and generalization ability are hard to achieve optimal. In this paper, we propose an algorithm to optimize the RBF neural network learning based on genetic algorithm; it uses hybrid encoding method, that is, encodes the network by binary encoding and encodes the weights by real encoding; the network architecture is self-adapted adjusted, and the weights are learned. Then, the network is further adjusted by pseudo inverse method or least mean square method. Experiments prove that the network gotten by this method has a better architecture and stronger classification ability, and the time of constructing the network artificially is saved. The algorithm is a self-adapted and intelligent learning algorithm.

This research studied how Amyotrophic Lateral Sclerosis patients can communicate after losing speech and typing abilities. To create a friendly and useful HCI system, this research studied the graphical user interfaces (GUI) through participant observations, to understanding how to innovate a better communication device for ALS patients and the elderly, to gaining a better quality of life.

Aiming at the large sample with high feature dimension, this paper proposes a back-propagation (BP) neural network algorithm based on factor analysis (FA) and cluster analysis (CA), which is combined with the principles of FA and CA, and the architecture of BP neural network. The new algorithm reduces the feature dimensionality of the initial data through FA to simplify the network architecture; then divides the samples into different sub-categories through CA, trains the network so as to improve the adaptability of the network. In application, it is first to classify the new samples, then using the corresponding network to predict. By an experiment, the new algorithm is significantly improved at the aspect of its prediction precision. In order to test and verify the validity of the new algorithm, we compare it with BP algorithms based on FA and CA.

Artificial Neural Networks (ANNs), as a nonlinear and adaptive information processing systems, play an important role in machine learning, artificial intelligence, and data mining. But the performance of ANNs is sensitive to the number of neurons, and chieving a better network performance and simplifying the network topology are two competing objectives. While Genetic Algorithms (GAs) is a kind of random search algorithm which simulates the nature selection and evolution, which has the advantages of good global search abilities and learning the approximate optimal solution without the gradient information of the error functions. This paper makes a brief survey on ANNs optimization with GAs. Firstly, the basic principles of ANNs and GAs are introduced, by analyzing the advantages and disadvantages of GAs and ANNs, the superiority of using GAs to optimize ANNs is expressed. Secondly, we make a brief survey on the basic theories and algorithms of optimizing the network weights, optimizing the network architecture and optimizing the learning rules, and make a discussion on the latest research progresses. At last, we make a prospect on the development trend of the theory.

Back-Propagation (BP) neural network, as one of the most mature and most widespread algorithms, has the ability of large scale computing and has unique advantages when dealing with nonlinear high dimensional data. But when we manipulate high dimensional data with BP neural network, many feature variables provide enough information, but too many network inputs go against designing of the hidden-layer of the network and take up plenty of storage space as well as computing time, and in the process interfere the convergence of the training network, even influence the the accuracy of recognition finally. Factor analysis (FA) is a multivariate analysis method which transforms many feature variables into few synthetic variables. Aiming at the characteristics that the samples processed have more feature variables, combining with the structure feature of BP neural network, a FA-BP neural network algorithm is proposed. Firstly we reduce the dimensionality of the feature factor using FA, and then regard the features reduced as the input of the BP neural network, carry on network training and simulation with low dimensional data that we get. This algorithm here can simplify the network structure, improve the velocity of convergence, and save the running time. Then we apply the new algorithm in the field of pest prediction to emulate. The results show that under the prediction precision is not reduced, the error of the prediction value is reduced by using the new algorithm, and therefore the algorithm is effective.

A back-propagation (BP) neural network has good self-learning, self-adapting and generalization ability, but it may easily get stuck in a local minimum, and has a poor rate of convergence. Therefore, a method to optimize a BP algorithm based on a genetic algorithm (GA) is proposed to speed the training of BP, and to overcome BP’s disadvantage of being easily stuck in a local minimum. The UCI data set is used here for experimental analysis and the experimental result shows that, compared with the BP algorithm and a method that only uses GA to learn the connection weights, our method that combines GA and BP to train the neural network works better; is less easily stuck in a local minimum; the trained network has a better generalization ability; and it has a good stabilization performance.

This paper proposes a radial basis function (RBF) neural network algorithm based on factor analysis (FA-RBF) with the architecture feature of RBF network when the data are high-dimensional and complex. By reducing the feature dimension of the original data, FA-RBF algorithm regards the data after dimension reduction as the inputs of the RBF network, and then trains and simulates the network. The algorithm obviously simplifies the network architecture. By analyzing an example, the results show when the algorithm's predicted precision is not reduced, the convergence velocity is improved, the running time is saved and the error of the predicted value is reduced. In order to test and verify the validity of the new algorithm, we compare it with the RBF neural network algorithm based on principal component analysis (PCA-RBF), the predicted results of FA-RBF algorithm are better than the results of RBF and PCA-RBF algorithm.

In this paper we discuss several progress of semi-supervised learning, making emphasis on semi-supervised classification. First, we introduce the history and basic concept and methods of semi-supervised learning, then explain the algorithms of semi-supervised learning in details, and discuss the current situation of semi-supervised learning, at last we look forward to the development trend of semi-supervised learning. © 2009 Binary Information Press.

Studing of insect ecology,predicting the pest occurrence,it does not require forecast the precise value,and it is difficult to get.So we only need forecast the pest occurrence trend.Fuzzy theory to deal with this problems has unique advantage.By using the technique of Fuzzy Regression,pointing at the data of 16 years,from 1989～2004,in Ningyang,Shandong province.The Fuzzy Regression model(FRM) is built to forecast the population dynamics of the corn aphid.Meanwhile the result of Fuzzy Regression model is contrasted to the result of multivariate linear regression model.The Fuzzy Regression model is proved to have higher veracity and better applicability.It's proved that the new model is verified.

The goal of statistical pattern feature extraction (SPFE) is `low loss dimension reduction'. As the key link of pattern recognition, dimension reduction has become the research hot spot and difficulty in the fields of pattern recognition, machine learning, data mining and so on. Pattern feature extraction is one of the most challenging research fields and has attracted the attention from many scholars. This paper summarily introduces the basic principle of SPFE, and discusses the latest progress of SPFE from the aspects such as classical statistical theories and their modifications, kernel-based methods, wavelet analysis and its modifications, algorithms integration and so on. At last we discuss the development trend of SPFE.

This paper roughly reviews the history of neural network, briefly introduces the principles, the features, and the applied fields of neural network, and puts emphasis on discussing the current situation of the latest research from parameters selection, algorithms improvement, network structure improvement and activation function; expounds the effect of neural network in modeling and forecast, from forward direction modeling and reverse direction modeling, describes the principles of modeling and forecast based on neural network in details, analyzes the basic steps of forecast using neural network, then discusses the latest research progress and the facing problems in this field, at last looks forward to the developing trend of this advancing front theory and its applied prospect in forecast.

As an important link of pattern recognition, pattern feature extraction and selection has been paid close attention by lots of scholars, and currently become one of the research hot spot in the field of pattern recognition. Its main purpose is "Low Loss Dimensionality Reduction"; it is generally divided into two parts, that is, linear pattern feature extraction and selection and nonlinear pattern feature extraction and selection. This article gives a general discussion of pattern feature extraction and selection, and introduces the frontier methods of this field, at last discusses the development tendency of pattern feature extraction and selection.

When we manipulate high dimensional data with Elman neural network, many characteristic variables provide enough information, but too many network inputs go against designing of the hidden-layer of the network and take up plenty of storage space as well as computing time, and in the process interfere the convergence of the training network, even influence the the accuracy of recognition finally. PCA,which is short for principal component analysis is a multivariate analysis method which transforms many characteristic variables to few synthetic variables. Not only can it eliminate relationship among characteristic variables but new variables can also hold most information of the original ones as well. In this paper we make full use of the advantages of PCA and the properties of Elman neural network structures to establish PCA-Elman based on PCA. The new algorithm reduces dimensions of the high dimensional data by PCA, and carry on network training and simulation with low dimensional data that we get, which obviously simplifies the network structure, and in the process, improves the training speed and generalization capacity of the Elman neural network. We prove the effectiveness of the new algorithm by case analysis. The algorithm improves the efficiency in network problems solving and is worth further generalizing.

The control effect of the natural enemy to cotton bollworm has temporal sequence and stage. The control is separated into four stages based on the evolvement of the cotton community dynamics,then the dynamics gray incidence degrees of the natural enemies to the cotton bollworm and the gray incidence quotieties of the natural enemies are obtained by using the gray incidence degree based on entropy.

Supervised learning is very important in machine learning. In this paper we discuss some progress of supervised learning. At first, we introduce the basic concept and methods of supervised learning; then explain several typical algorithms of supervised learning in details, the algorithms covered are Bayesian networks, decision tree, k-nearest neighbor, supervised manifold learning and support vector machines; at last we point out several developing directions of supervised learning.

The appearance of pest is a nonlinear system.There are many predicting factors that influence the appearance of pest.They are interrelated to some extent.When using neural network to predict,it is not propitious for the design and computation.combined the principle of factor analysis and neural network and builds the model basing on the combination of factor analysis and neural network were built.By factor analysis the dimensionality of predicting factors was reduced,and the data after dimension reduction was regarded as the input of the network,then the predicting results after training.By analyzing the second period prediction of Helicoverpa armigera(Hubner) of Yuncheng,Shandong,it is proved that the prediction inaccuracy of the new model is not reduced,the convergence velocity speed up,and the error of prediction value is reduced.It shows that this model has wide application prospects in the aspects of the prediction of plant diseases and insect pests.