A Study on Method of Normalizing the Data of Material Corrosion

摘要：

Man-made neural networks have been more used on analyze corrosion data in recent years. Original corrosion data come from accelerated corrosion test, outfield exposed corrosion test, or practical applied occasion, and they are some data groups with many dimensions and include information involving material, corrosion rate, corrosion medium, environment factors, load condition, and so on. Original corrosion data must be normalized before analyzing by man-made neural networks. For the same data, if the method to normalize data is different, the analyzing result might be very different and this might affect objectivity of analyzing result. So it is necessary to standardize the method of normalizing corrosion data. In this paper, the method to normalize corrosion data is discussed, and an applied example is brought forth, which can be used for reference.

关键词： Neural Network Material Corrosion Data Normalization

作者: Desheng Han Xinru Jia

作者单位: College of Chemistry and Molecular Engineering, Peking University, Beijing, 100871, China Guangdong College of Chemistry and Molecular Engineering, Peking University, Beijing, 100871, China

会议类型: 国际会议

会议名称: The 2nd Asian Materials Database Symposium(第二届亚洲材料数据库会议 AMDS 2010)

会议地点: 三亚

会议语种:英文

页码: 318-322

在线出版日期: 2010-03-10（万方平台首次上网日期，不代表论文的发表时间）