Abstract: Ensemble classification can effectively improve the classification performance by combining several weak classifiers according to a certain rule. In the process of combination, the importance of each weak classifier to the classification result is often different. Extreme learning machine is a new learning algorithm for training single-hidden-layer feedforward neural networks. In this paper, the extreme learning machine is selected as a base classifier, and a weighted ensemble method based on differential evolution is proposed. The proposed method optimizes the weights of each base classifier in the ensemble method based on differential evolution algorithm. Experimental results show that the proposed method has higher classification accuracy and better generalization ability compared with the simple-voting method and Adaboost ensemble method.

Key words: differential evolution, extreme learning machine, ensemble classification, weighting

摘要： 集成分类通过将若干个弱分类器依据某种规则进行组合，能有效改善分类性能。在组合过程中，各个弱分类器对分类结果的重要程度往往不一样。极限学习机是最近提出的一个新的训练单隐层前馈神经网络的学习算法。以极限学习机为基分类器，提出了一个基于差分进化的极限学习机加权集成方法。提出的方法通过差分进化算法来优化集成方法中各个基分类器的权值。实验结果表明，该方法与基于简单投票集成方法和基于Adaboost集成方法相比，具有较高的分类准确性和较好的泛化能力。

关键词: 差分进化, 极限学习机, 集成分类, 加权