径向基-偏最小二乘-贝叶斯方法及其在化学模式分类中的应用

提出一种用于模式分类的RBF-PLS—Bayes方法。它集成地应用径向基(RBF)变换与偏最小二乘(PLS)方法，从原有模式中提取出分类能力甚强的成分，然后进行贝叶斯(Bayes)判别。这种集成方法尤其适用于复杂化学信息的模式分类，本文将其应用于两种类型的化学模式分类问题，均取得了令人满意的效果。与经典的判别分析方法和单纯的神经网络方法相比，具有明显的优越性。

Radial Basis Function-Partial Least Square-Bayes Method and Its Application in Chemical Pattern Classification

A novel method called radial basis function-partial least square-Bayes (RBF-PLS-Bayes) algorithm is proposed which can be applied in pattern classification. This method combines the RBF transformation with the PLS method, and selects the components whose classification ability is relatively high in order to carry out the bayes discrimination. Such a combined method is especially applicable to the pattern classification of complex chemical information. By applying the method in two kinds of chemical pattern classification problems we get a satisfying result. Compared with the traditional discriminant analysis method and neural networks method, this method has a distinct superiority.