| 一种新的最大间隔分类器的求解算法及其在测井岩性识别中的应用 |
| |
| 引用本文： | 罗明璋,张翔,刘振丙.一种新的最大间隔分类器的求解算法及其在测井岩性识别中的应用[J].石油天然气学报,2012(2):98-102. |
| |
| 作者姓名： | 罗明璋 张翔 刘振丙 |
| |
| 作者单位： | [1]长江大学电子信息学院；长江大学地球物理与石油资源学院,湖北荆州434023 [2]长江大学地球物理与石油资源学院,湖北荆州434023 [3]桂林电子科技大学电子工程与自动化学院,广西桂林541004 |
| |
| 基金项目： | 国家自然科学基金项目（40727001;40774073）; 国家“973”规划项目（2007CB209607） |
| |
| 摘 要： | 首先介绍了尺度不变凸壳的概念,提出当2个凸壳独立时,可以通过求解2个凸壳的最大间隔来建立一种新的分类器。然后通过论证尺度不变凸壳与最小闭球的关系,提出了利用最小闭球求解分类器的快速算法。通过对测井岩性识别的试验表明,与已有的方法相比,在取得相同分类成功率的同时,该算法的核估计次数和训练时间明显减少。
|
| 关 键 词： | 尺度不变凸壳 最大间隔 最小闭球 岩性识别 |
A Novel Algorithm for Maximal Margin Classifier and Its Application in Logging Lithologic Identification |
| |
| LUO Ming-zhang,ZHANG Xiang,LIU Zhen-Bing.A Novel Algorithm for Maximal Margin Classifier and Its Application in Logging Lithologic Identification[J].Journal of Oil and Gas Technology,2012(2):98-102. |
| |
| Authors: | LUO Ming-zhang ZHANG Xiang LIU Zhen-Bing |
| |
| Affiliation: | (College of Geophysics and Oil Resources,Yangtze University,Jingzhou 434023,Hubei,China) |
| |
| Abstract: | By introducing the notion of"scaled convex hull"(SCH) generated by the training points,a novel classifier could be constructed by maximizing the margin between two SCH when they were independent.Then,a fast algorithm to solve the classifier is presented by building the relationship between the SCH and the minimum enclosing ball.The experiments on the data of logging lithological character identification show that the proposed method may achieve better performance than the state-of-the-art methods,in terms of kernel evaluations and execution time. |
| |
| Keywords: | scaled convex hull maximal margin the minimum enclosing ball logging lithologic character identification |
| 本文献已被 CNKI 维普 等数据库收录！ |
