Fuzzy Prediction for Time-Series Data-A Case Study at Taichung City Open Data of Air Pollution

摘要

In this big data era, we always use the statistical or mathematical model for historical data analysis, and the use of data analysis results to predict future changes and trends. In 1965. Zadeh's fuzzy theory appeared in order to analyze and discuss the fuzzy number. At present, the fuzzy theory has been widely used in many fields. In recent years, the urban air quality is deteriorated, the impact of air pollution on human health and the air quality is increasingly valued. So to establish the air pollution data prediction model is in order to let people know the change of air pollution, and we will have time to take appropriate strategies to reduce the negative impact of air pollution on human health. In the research, I selected the total of eleven years of air pollution monitoring data between 2005 and 2015 from the Taichung air monitoring station. Using the method of autoregressive model to establish the time series prediction model of air pollution and to simulate the trend of air pollution in Taichung City. The result shows that probability statistics model can effectively predict the changes of air pollution from the data, and be able to provide changes in air pollution trends. Therefore, people or the government can do some precautionary measures in advance to reduce health hazards.