摘要
针对高维数据中存在冗余以及极限学习机(ELM)存在随机给定权值导致算法性能不稳定等问题,将限制玻尔兹曼机(RBM)与ELM相结合提出了基于限制玻尔兹曼机优化的极限学习机算法(RBM-ELM).通过限制玻尔兹曼机对原始数据进行特征降维的同时,得到ELM输入层权值和隐含层偏置的优化参数.实验结果表明,相比较随机森林,逻辑回归,支持向量机和极限学习机四种机器学习算法,RBM-ELM算法能获得较高的分类精度.
Since the high-dimensional data has the redundancy problems and the extreme learning machine（ELM） has the problem of the instability which caused by setting the input weights and bais randomly,this paper proposed an improved algorithm of extreme learning machine based on restricted boltzmann machine（RBM-ELM）.RBM is used to optimize the weights of input layer and the bias of hidden layer,meanwhile to extract discriminative lowdimensional features from the raw data.The experimental results show that compared with the random forest,logistic regression,support vector machine（SVM） and ELM,RBM-ELM algorithm can achieve higher classification accuracy.
出处
《数学的实践与认识》
北大核心
2016年第11期157-161,共5页
Mathematics in Practice and Theory
基金
国家自然科学基金(61272315
60842009)
浙江省自然科学基金(LY12H29012
Y1110342)
关键词
限制玻尔兹曼机
极限学习机
降维
输入层权值
隐含层偏置
restricted boltzmann machine
extreme learning machine
dimensionality reduction
input layer weights
bias of hidden layer
