微量元素与乳腺癌死亡率的模式识别研究

Research on pattern recognition for the mortality of breast cancer and trace element

采用基于统计学习理论的支持向量机方法,建立了硒、铜、锌、镉、铬、锰和砷7种微量元素的平均摄入量与每10万人中的乳腺癌死亡人数的预测模型,并对6个国家或地区进行了预测,还与传统的人工神经网络方法进行了比较。结果表明:在根据微量元素摄入量进行乳腺癌发病率预测方面,支持向量机方法有明显的优势。

By using support vector machine method based on the statistical study theory, This paper has established the predicting model of breast cancer mortality in ten thousand persons and average gain quantity of seven kind of trace elements - selenium, copper, zinc, cadmium, chromium, manganese and arsenic, and has carried on the forecast to 6 countries or areas, but also has carried on the comparison with the traditional artificial neural network method. The result indicated that support vector machine method has the obvious superiority in the forecast of breast cancer incidence rate based on gain quantity of trace element.