Study Of Neural Network Ensemble Algorithm And Application Of It In Gene Expression Data Analysis

Neural network ensemble(NNE) is a hot topic of neural compute,which has been applied in many fields.A neural network ensemble is a very successful technique where the outputs of a set of separately trained neural network are combined to form a unified prediction.Negative correlation learning(NCL) algorithm for training NNE is to encourage different individual networks in an ensemble to learn different parts or aspects of a training data so that the ensemble can learn the whole training data better.NCL can create negatively correlated neural networks using a correlation penalty term in the error function.Based on microarray experiment, the expression level of thousands of genes can be simultaneously observed, and the method of the analysis for the gene expression datas is hot in bioinformatics. The Datas have some traits,such as small samples, high dimensionality,non linearity, too.The theory and method of neural network ensemble were studied in the given gene expression data. Many methods have been discussed. A new algorithm based on NCL has been put forward in slow speed of training. Finally, a software—GeneNCL based NCL for gene expression data is achieved and a classifier model is put forward: the gene subset are extracted by the method of signal to noise ratio ,and datas are normalized by min-max method,a classifier is built by the method of NCL, a classifier model based on above theories is successfully applied in a typical gene expression datas.