基于线性回归模型的颜色与物质浓度辨识应用研究

Study of color and material concentration identification based on linear regression model

以2017年全国大学生数学建模竞赛C题为主题,运用多元线性回归模型对5种物质的颜色读数与物质浓度之间的关系进行研究,引入绝对误差和相对误差的概念对模型进行误差分析。最后,通过4组实验探讨数据量增加与减少、颜色维度增加与减少对模型的影响。

The relationship between the color reading of five substances and the material concentration was studied with a multivariate linear regression model for question C of the 2017 national university students'mathematical modeling competition.The concept of absolute error and relative error is introduced to analyze the model error.Four groups of experiment were conducted to study the influence on the model from increases and decreases of data quantity or color dimensions.