地学前缘 ›› 2019, Vol. 26 ›› Issue (4): 125-130.DOI: 10.13745/j.esf.sf.2019.5.19

• 大数据算法研究 • 上一篇    下一篇

关联规则算法在粤西庞西垌地区元素异常组合研究中的应用

刘心怡，周永章   

  1. 1. 中山大学 地球环境与地球资源研究中心, 广东 广州 510275
    2. 中山大学 地球科学与工程学院, 广东 广州 510275
    3. 广东省地质过程与矿产资源探查重点实验室, 广东 广州 510275
  • 收稿日期:2018-07-30 修回日期:2019-05-19 出版日期:2019-07-25 发布日期:2019-07-25
  • 通讯作者: 周永章(1963—)，男，教授，博士生导师，主要从事大数据、数学地质与地球化学研究。
  • 作者简介:刘心怡(1994—)，女，硕士研究生，大数据挖掘、机器学习与地球化学研究方向。
  • 基金资助:
    国家重点研发计划项目(2016YFC0600506)；国家自然科学基金项目(41273040)；中国地质调查局地质调查项目(12120113067600)

Application of association rule algorithm in studying abnormal elemental associations in the Pangxidong area in western Guangdong Province, China

LIU Xinyi,ZHOU Yongzhang   

  1. 1. Center for Earth Environment & Resources, Sun Yat-sen University, Guangzhou 510275, China
    2. School of Earth Sciences & Engineering, Sun Yat-sen University, Guangzhou 510275, China
    3. Guangdong Provincial Key Laboratory of Mineral Resources and Geological Processes, Guangzhou 510275, China
  • Received:2018-07-30 Revised:2019-05-19 Online:2019-07-25 Published:2019-07-25
  • Supported by:
     

PDF (PC)

249

摘要： 以粤西庞西垌矿床远景区1∶5万水系沉积物地球化学测量及异常查证数据为基础，应用Python编程语言开展关联规则算法的应用案例研究。结果显示，Apriori算法可以有效挖掘元素组合之间的关联规则数据集。当Au、Cu、Sb在异常值范围内时，出现As为异常的可能性是95.5%。Apriori算法挖掘的关联规则符合实际，组合异常的强规则与研究区已知矿化地段的异常组合重合性较高。可以推论，面对海量的地球化学数据,逐个进行元素分析较为耗时而且无法观测到元素之间的关系，通过关联规则算法找出元素异常组合规律的办法，使之最大限度地保留元素之间的相关信息，可以用来寻找隐藏的元素组合以及其中的潜在相关性。未来构建指示找矿的成矿关联规则数据库并进行矿床预测，将比运用传统的方法更加便捷。

 

关键词: 地质大数据, 大数据挖掘, 关联规则算法, Apriori算法, 元素异常组合

Abstract: We conducted a case study on the application of association rule algorithm (programmed in Python) using the original 1∶50000 geochemical survey and anomaly verification data of stream sediments in the Pangxidong deposit prospect district in the southern section of the Qinzhou Bay-Hangzhou Bay orogenic belt. The results showed that the Apriori algorithm can effectively mine the association rule itemsets of elemental combinations. For example, we found that As had a 95.5% probability being abnormal when Au, Cu and Sb in the itemset were in abnormal range. The association rules selected by Apriori algorithm were in line with survey results; and the strong rules of combination anomalies had high agreement with the anomaly combinations of known mineral deposits in the study area. Facing with massive geochemical survey data, it is often time-consuming to understand elements one by one, and in many cases it is impossible to observe the relationship among them. Therefore, it is advantageous to expose the abnormal combination rules of elements using association rule algorithm. By doing so, related information among various elements can be stored to a great extent and used to find the hidden combinations of elements and potential correlations among them. And compared to traditional methods, it can be more convenient and effective to establish data bases of metallogenic association rules and to carry out mineral deposit prediction.

Key words: geological big data, big data mining, association rules, Apriori algorithm, abnormal element association

中图分类号: