A CASE STUDY OF ENVIRONMENTAL MONITORING DATA ANALYSIS AND FORECASTING MODEL

摘要

Environmental related data are increasing rapidly in China and have been used in many fields based on data processing and mining methods or technology. Through a recent case study of Beijing, the paper aims to provide a better forecasting result by optimizing existed forecasting model by using multiple linear regression method. Firstly, the collection and processing of air quality monitoring data are introduced, and then focusing on traditional multiple linear regression we propose two optimization methods: "input-variable- based optimization" and "stepwise-regression-based optimization", along with correlation coefficient test, F-test and t-test by SPSS. Finally, we evaluate how well the optimized forecasting model is working from two aspects: accuracy and usefulness.