支持向量机算法研究土壤微量元素含量与白血病的关系

Support vector machine applied to study relationship between leukocythemia and soil trace element contents

应用计算机支持向量机算法处理我国29个省、市、自治区白血病死亡率的统计数据和土壤中微量元素含量的对应关系,得到明显的复相关关系。可以认为：砷(As)等元素含量与白血病死亡率负相关,汞(Hg)等元素则与白血病死亡率正相关。已知大量砷能致癌,但近年来发现少量氧化砷治疗白血病效果显著,其原因可能是微量砷能导致癌细胞凋亡,土壤中的微量砷和白血病死亡率的负相关似可据此解释。已知元素硒(Se)有明显的抗癌效果,而Hg对Se有拮抗作用。因此Hg含量对白血病死亡率的正相关也是可以理解的。

Support vector machine algorithm has been applied to find the correlation between the mortality of leucocythemia and the contents of trace elements in the soils of 29 provinces, cities and autonomous regions of China. It has been found that there is some obvious multiple correlation between the leucocythemia mortalities and the soil contents of arsenic （As）, mercury （Hg） and other trace elements. The As contents and the leucocythemia mortalities exhibit negative correlation. This fact can be explained by the apoptotic effect of cancerous cells by trace amount arsenic trioxide （As2O3）. While the correlation between Hg content and leucocythemia mortality can he explained by the antagonistic action of Hg to cancer-prevention effect of Se.