空气质量数据校准的统计学习与建模

Statistical Learning and Modeling of Calibration for Air Quality Data

为2019年"高教社杯"全国大学生数学建模竞赛D题"空气质量数据的校准"给出了可行的解法,根据赛题所给的数据,对自建点数据和国控点数据进行了探索性数据分析,分析了自建点数据和国控点数据存在差异的原因,并采用多元线性回归模型和前馈神经网络模型对自建点数据进行统计学习和校准,并对学生在参赛中出现的方法进行了简要的说明与点评.

This paper provides a feasible solution for problem D"Calibration of Air Quality Data"in 2019 Higher Education Press Cup China Undergraduate Mathematical Contest in Modeling.Based on the data given in the problem,exploratory data analysis as well as correlation and difference data analysis on the data from the self-built station and the national monitoring station are given.Furthermore,a multiple linear regression model and BP neural network model are built respectively to make statistical learning and calibrate the data of self-built station,and brief description and comments for the methods of building mathematical model by students in the contest paper is also given.