Abstract: Extreme learning machine (ELM) is a simple and effective SLFNs (single hidden layer feedforward neural networks) learning algorithm, in recent years has become one of the hot areas of machine learning research. But single hidden layer node lacks judgment to some extent. The classification accuracy depends on the number of hidden layer nodes. In order to improve the sense ability of single hidden layer node, the support vector machine (SVM) is combined with ELM, a simplified SVM-ELM model. At the same time, in order to avoid the subjectivity of human to choose parameters, using particle swarm optimization (PSO) algorithm to automatically select the parameters, finally PSO-SVM-ELM model is established. Experiments show that classification accuracy of the model is improved than the SVM-ELM and ELM, and the model also has good robustness and generalization.

Key words: particle swarm optimization (PSO), support vector machine (SVM), extreme learning machine (ELM), SVM-ELM

摘要： 极限学习机（extreme learning machine，ELM）是一种简单易用、有效的单隐层前馈神经网络（single hidden layer feedforward neural networks，SLFNs）学习算法，近几年来已成为机器学习研究的热门领域之一。但是ELM单个隐层节点的判断能力不足，分类正确率的高低在一定程度上取决于隐层节点数。为了提高ELM单个隐层节点的判断能力，将支持向量机（support vector machine，SVM）和ELM结合，建立一种精简的SVM-ELM模型。同时，该模型为了避免人为选择参数的主观性，利用粒子群算法（particle swarm optimization，PSO）的全局搜索最优解对参数进行自动优化选取，建立了PSO-SVM-ELM模型。实验证明，该模型较SVM-ELM和ELM分类精度有较大的提高，具有很好的稳健性和泛化性。

关键词: 粒子群算法（PSO）, 支持向量机（SVM）, 极速学习机（ELM）, SVM-ELM