| Summary: | 碩士 === 國立勤益科技大學 === 工業工程與管理系 === 101 === The science and technology progression of human being led to the development of transportation, making people more convenient who lived in cities. Transportation has become important tool in our daily life, but accident would also become a big problem that we must solve. There were many reasons that cased accident such as different environment, different times, different places, human factors and road conditions. Therefore, the improvement of traffic safety that made people safer, reducing the frequency of accident and the loss of life and money of people was government's responsibilities. This research will use data mining skills in traffic safety area, and make people on the road safer as purpose. We use accident records from 2008 to 2011 as data source, investigating the relevance among casualty, vehicle machinery, weather conditions and road design. This research will use classification, CART and K-man doing feature selection and clustering for accident database respectively. Then, we will use Support Vector Machines, Back Propagation Neural Network and Bayesian Network analyzing this new database. Comparing the accuracy which assesses the performance of classification is applied good way to solve different situation database. This research will become a research reference for related researchers and police officers.
|
|---|
