Taiwan Passenger Transportation Industry's Safety Evaluation Mechanism: Using Machine Learning Based Approach

• Main Authors: Ching-Hao Chen, 陳慶豪
• Other Authors: Yu-Chung Tsao
• Format: Others
• Language: en_US
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/43q63z

碩士 === 國立臺灣科技大學 === 工業管理系 === 106 === Vehicle crash is one of the most common reason that caused deaths and injuries all around the world. Also, it would make great loss. In 2017, there are two serious vehicle accidents happened in Taiwan, Nangang Tour bus accident and Aloha coach crash. These two accidents caused 39 killed and 22 injured in total. After these happened, the safety consciousness rises and the government has duty to provide safe environment for all people. However, there is no proper safety evaluation mechanism in Taiwan nowadays. In this study, we combine CSA program of FMCSA in the US, Renting Bus safety evaluation in Japan, and some of the service evaluation in Taiwan to develop a new passenger industry’s safety evaluation mechanism which could be suitable for Taiwan. In this study, we designed a two-stage decision making approach to develop this mechanism. We apply fuzzy AHP to set up the initial safety rank of each company in the first stage. In the second stage we use machine learning approaches to analyze the data and to adjust the attributes’ weights. We can calculate each company’s safety performance and divide them into different safety ranks. These ranks can be an index when we need service and can be used in insurance industry to do differential pricing. Also, we can get some insights from the importance of each attributes.