一种过程支持向量机模型及其在油水层判别中的应用

Process support vector machine model and its application to the discrimination of oil and water layers

针对时变信号动态模式分类问题,提出一种过程支持向量机模型PSVM.过程支持向量机的输入可为时变函数(或函数向量),通过核函数变换将动态模式映射到高维特征空间,经过对训练样本集中过程支持向量类别特性的学习,可直接建立时变信号的分类模型.给出过程支持向量机的一般模型和求解算法,油水层判别仿真实验结果验证了模型和算法的有效性.

Aiming at the problem of classification of Time-Varying Signal patterns, this paper presents a Process Support Vector Machine (PSVM) model. The input of PSVM can be functions with time-varying (or function vector) according to kernel function transforming, map dynamic pattern to high-dimensional feature space. For the study of PSVM classification in training sample set, it can build classification model of time-varying signals directly. The model of PSVM and its algorithm for solving were given. The results of simulation experiments confirmed the efficiency of the model and algorithm.