摘要
Ｂｏｏｓｔｉｎｇ是一种新型的机器学习算法 ,其主要用于提高回归算法的性能。介绍了一种以ＲＢＦ神经网络为基础学习机的Ｂｏｏｓｔｉｎｇ回归算法 ,并将此算法应用于油品辛烷值分析中 ,与常用的油品分析技术偏最小二乘法 (ＰＬＳ)、多元线性回归(ＭＬＲ)方法和单个ＲＢＦ神经网络的拟和预测效果对比分析。结果显示 ,该算法具有学习速度快、跟踪性能好、范化能力等优点。
Boosting is a new machine learning algorithm,which is usually utilized to improve the performance of an ensemble of regression algorithms.A boosting regression algorithm using RBF neural networks as a base learner is proposed.Meanwhile this algorithm is applied in the prediction of gasoline octane number.In comparison with other traditional techniques,such as PLS method,MLR method and single RBF neural network,it shows that this method features high learning speed,good approximation and excellent generalization ability.
出处
《石油化工自动化》
CAS
2004年第1期31-33,50,共4页
Automation in Petro-chemical Industry
