Cross-city crash severity analysis with cost-sensitive transfer learning algorithm
作者：
Highlights：
• Propose unified feature representation for cross-city crash records feature alignment.
• Develop two transfer learning based cross-city crash severity models.
• Adopt the cost-sensitive learning method to tackle the class imbalance issue.
• Identify more accurate significant crash contributing factors to crash severity level.
• Extract and explain common crash contributing factors of target and source cities.
摘要
•Propose unified feature representation for cross-city crash records feature alignment.•Develop two transfer learning based cross-city crash severity models.•Adopt the cost-sensitive learning method to tackle the class imbalance issue.•Identify more accurate significant crash contributing factors to crash severity level.•Extract and explain common crash contributing factors of target and source cities.
论文关键词：Crash severity analysis,Traffic safety,Cross-city,Transfer learning,Cost-sensitive learning,Imbalanced dataset
论文评审过程：Received 17 August 2021, Revised 5 May 2022, Accepted 8 July 2022, Available online 19 July 2022, Version of Record 25 July 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2022.118129
