摘要
贝叶斯网络是一种建立在概率和统计理论基础上的数据分析和辅助决策工具,利用其构造出的树扩展朴素贝叶斯网络分类器是目前最优秀的分类器之一。针对石油勘探中测井数据的特殊性,利用贝叶斯网络预测出其对应的岩性,并介绍了使用此方法进行岩性预测的算法过程。通过BNT软件包用Matlab语言构建了分类器,并由实验结果的分析说明了此分类器的优点。
Bayesian Networks is a kind of data analyzing and assistant decision-making tool which is based on probability and statistic thoery.The Tree Augmented Naive Bayesian Classifier(TANC) is one of the best Bayesian networks classifiers,which is constructed with BN.To the question of particularity of well data in oil exploration,we use BN to identify the corresponding lithologies ,and discusses the arithmetic progress of lithologies identification.The classifier is constructed using Matlab based BNT software package.And the experiment result illuminates the advantage of this classifier.
出处
《微计算机信息》
北大核心
2006年第09S期284-286,共3页
Control & Automation
基金
国家科技部十五攻关项目2001BA605A-08-05