本文在加权径向基函数这一统一方法的框架下,对常用的几种软计算方法(包括神经网络、小波网络、模糊系统、贝叶斯分类器、模糊划分)以及支撑矢量机等机器学习方法的内在机理做了进一步的研究.特别地,对于支撑矢量机这种新的学习方法,文中分析了它和神经网络等方法之间的异同处,并尝试性地把其纳入统一的框架下,从而为支撑矢量机及软计算方法的研究与应用提供了理论上的指导. In this paper, under the framework of the uniform method of weighted radial basis function, several commonly used soft computing methods (including neural networks, wavelet networks, fuzzy systems, Bayesian classifiers, fuzzy partitioning) and machine learning such as support vector machine Method of the internal mechanism to do further research.Especially, for the support vector machine this new learning method, the article analyzes the similarities and differences between it and the neural network and other methods, and tentatively put it into a unified framework , Which provides theoretical guidance for the research and application of support vector machines and soft computing methods.