探究天氣因素對市區交通事故之影響

交通事故對於城市的影響，並不只於當事人而已，經常會對其他道路使用者造成不便。而交通事故發生的原因複雜，例如道路狀況、交通流量、號誌管制、及駕駛人本身因素等，而天候也可能是交通事故的因素。隨著機器學習的技術蓬勃發展，全球多數政府對資料的高透明度，讓巨量資料分析的運用日趨生活化。本研究藉由天氣因素中的風向、風速、雨量、氣壓、氣溫、紫外線等，運用貝氏分類器、決策樹、近鄰演算法、隨機森林機等學習工具，分析天候與市區交通事故之間可能的關聯。

關鍵字

天氣因素；交通事故；機器學習

並列摘要

The impact of traffic accidents is not limited to the parties involved, and often causes inconvenience to other road users. The causes of traffic accidents are complex, such as road conditions, traffic flow, signal control, and drivers themselves. Weather may also be a factor in traffic accidents. With the vigorous development of machine learning technology and the high transparency of data by governments around the world, the application of big data analysis has become more closer to people's daily life. This study uses machine learning methods such as Decision Trees, K-Nearest Neighbors, Random Forest, and Bayesian classifiers to examine the role of weather conditions such as the wind direction, wind speed, rainfall, air pressure, temperature, ultraviolet rays, etc. on traffic accidents.

並列關鍵字

Weather ； Traffic Accidents ； Machine Learning

主題瀏覽