摘要
传统的交通流量预测算法没有考虑到商圈的性质、交通拥堵等情况,而这些影响因子在智能交通时代对于人们的出行有着举足轻重的影响。同时,其预测目标也不是起点和终点的流量。经过对流量预测算法进行研究,给出基于GBRT的集成学习模型对交通领域的起始地-目的地对的流量进行预测,并利用平均绝对误差(MAE)和均方根误差(RMSE)评价指标对模型进行评估,结果证明使用GBRT进行流量预测具有较好的效果。
The traditional traffic flow prediction algorithm does not consider the nature of the business circle, traffic congestion, etc., and these impact factors have a significant impact on people's travel in the era of intelligent transportation. At the same time, its predicted goal is not the flow of the origin and destination. After research of the traffic prediction algorithm, presents an integrated learning model based on GBRT to forecast the flow of given origin and destination in the traffic field, and evaluates the model by Root Mean Square Error (RMSE), Mean Absolute Error(MAE). The results show that GBRT has a good effect in traffic prediction.
作者
周鑫
ZHOU Xin(College of Computer Science, Sichuan University, Chengdu 610065)
出处
《现代计算机》
2019年第7期38-41,共4页
Modern Computer
作者简介
周鑫(1993-),男,四川成都人,硕士,研究方向为数据挖掘、城市计算.
