基于邻域粗糙集的极速学习机算法

摘要

The extreme-learning machine (ELM) algorithm is a new learning algorithm for the single hidden layer feedforward neural networks (SLFNs).Combining neighborhood rough set theory with extreme-learning machine,we propose a new algorithm of extreme -learning machine based on neighborhood rough set theory.Firsdy,attribute reduction is done using the neighborhood rough set theory,and then the reduced datasets are trained and predicted using ELM.Final experimental results demonstrate that the proposed algorithm has higher training accuracy and prediction accuracy compared with the traditional ELM.%将基于单隐层前馈神经网络(SLFN)提出的极速学习机(ELM)算法和邻域粗糙集理论进行结合,提出基于邻域粗糙集的极速学习机算法,采用邻域粗糙集对样本集进行属性约简,去掉冗余属性,利用ELM对约简后的数据集进行学习,并对数据样本进行预测.实验表明ELM算法相比具有更高的训练精度和测试精度.