Neural Network Ensemble Based On Boosting For Classification

Neural network ensemble can significantly improve the generalization ability of learning systems through training a finite number of neural networks and then combining their results. It is not only helpful for scientists to investigate machine learning and neural computing but also helpful for common engineers to solve real-world problems using neural network techniques. Therefore neural network ensemble has been regarded as an engineering neural computing technology that has great application prospect. Also it has become a hot topic in both machine learning and neural computing communities. Boosting is a method for improving the accuracy of any given learning algorithm, which generate multiple versions of a hypothesis and combine them to create an aggregate hypothesis. Based on the detailed analysis of the adaboost algorithm, this paper presents two new adaboost algorithms which are the method based on optimizing training dataset and the method based on optimizing the weights of weak classifiers using LMS. The theories and experimental results prove that the two new algorithms have less training error and better generalization ability. In this paper, we use adaboost algorithm to generate single neural network for neural network ensemble, then do gene pattern recognition. The experimental results have verified the feasibility and validity of neural network ensemble based on Boosting for gene classification which often has few samples and high dimension.