| Summary: | 碩士 === 輔仁大學 === 資訊管理學系碩士班 === 106 === There is a well-developed transportation network in Taiwan and the number of vehicles is on the rise. As a result, the number of accidents caused by traffic incidents has also risen. According to the statistics of the cause of death in the year 2015 published by the Ministry of Health and Welfare of the Republic of China, the accidental injuries are ranked No. 6 among top ten injuries. In recent years, the government in Taiwan has actively conducted a large amount of data analysis and research, but there is no study showing the impact of vehicles in fatal traffic accidents at present. The purpose of this paper is to identify the relationship between the vehicles and fatal car accidents and build a traffic accidents prediction model using data mining technology. The data used in this study was compiled by the National Highway Traffic Safety Administration in American which contains traffic accidents occurred in 2015. One challenge is that the available data are complex and may contain noise. This raises the question of how to select the most important explanatory factors to achieve an acceptable performance and accuracy of prediction process. This research suggests to principal component analysis, which is widely used to extract key factors from the data set. A prediction model is built by compare back-propagation neural networks and logistic regression in this study. The results of the prediction model are investigated to gain insights of the reasons which caused accidents. An analysis is made after the results of the model is validated to see if the traffic prediction model can be applicable in Taiwan.
|
|---|
