| 形态相关系数及其在地球化学数据分析中的应用 |
| |
| 引用本文： | 余先川,张冠鹏,姚旺,等.形态相关系数及其在地球化学数据分析中的应用[J].地质学刊,2019,43(1):103-110. |
| |
| 作者姓名： | 余先川 张冠鹏 姚旺 等 |
| |
| 作者单位： | 北京师范大学信息科学与技术学院 |
| |
| 基金项目： | 国家自然科学基金项目“基于稀疏成分分析的找矿信息识别”（41672323, 41272359），国土资源部公益基金项目“地质大数据综合分析关键技术研究”（201511079-02) |
| |
| 摘 要： | 地球化学数据元素间的相关分析具有重要意义。地球化学数据是一种不遵循正态分布的成分数据，其封闭特征存在挖掘和分析的困难和障碍，因此许多传统的统计方法不适合使用。主要通过元素值的趋势分析了地球化学元素之间的相关关系，并提出了形态相关系数的概念。该方法不需要数据服从正态分布，可以忽略闭包特征的影响。实验表明，该方法简单、稳定、准确，可以显示数据元素之间的关系。此外，方法计算过程消除了回溯，因此适用于大数据的实时和动态分析。
|
| 关 键 词： | 形态相关系数 地球化学数据 成分数据 |
Morphological correlation coefficient and its application in geochemical data analysis |
| |
| Yu Xianchuan,Zhang Guanpeng,Yao Wang,et al.Morphological correlation coefficient and its application in geochemical data analysis[J].Jiangsu Geology,2019,43(1):103-110. |
| |
| Authors: | Yu Xianchuan Zhang Guanpeng Yao Wang |
| |
| Abstract: | Correlation analyses are important between geochemical data elements. Geochemical data is a type of compositional data which does not obey the normal distribution and its closure feature present mining and analysis difficulties and obstacles; therefore, many traditional statistical methods are not suitable for use. This paper mainly analyzed the correlation relationship between the geochemical elements through the elemental values trend and proposed the Morphological Correlation Coefficient concept. This method didn,t need the data obey the normal distribution and can ignore the influence of closure feature. Experiments show that the method is simple, stable and accurate to show the relation between data elements. In addition, the method calculation process eliminates backtracking, so it is appropriate in real-time and dynamic analyses of large data. |
| |
| Keywords: | morphological correlation coefficient geochemical data compositional data |
|
| 点击此处可从《地质学刊》浏览原始摘要信息 |
| 点击此处可从《地质学刊》下载免费的PDF全文 |
