부산대학교 도서관

As advanced automotive technology continues to reshape the transportation landscape worldwide; automobiles have emerged as the preferred mode of commute. While this transformation has brought unprecedented convenience to road travel, it has also given rise to a pressing issue: car accidents. This paper introduces a predictive method designed to assess the likelihood of such accidents by analyzing the key contributing factors. The approach involves the extraction of relevant features through the development of a dedicated system, which is then utilized to generate binary accident probability predictions. Notably, the dataset and machine learning algorithm employed yield impressive results, achieving an accuracy rate of 98%. This tool holds significant promise for addressing safety concerns within the self-automation automobile industry in the future. The outcomes underscore the method's value to various stakeholders, including car drivers, automotive safety professionals, and companies involved in the sector.