基于多源数据融合的城市出租车载客出行特征研究——以岳阳市为例

为探究城市出租车载客出行特征，在出租车GPS轨迹大数据基础上，融合居民出行调查数据、城市土地利用数据及天气数据，构建出租车载客出行量回归模型，得出出租车载客出行量与片区岗位数、天气状况、时段、片区面积有较强的相关性，而基于RBF神经网络构建的回归模型在上述4个因素的基础上增加了片区常住人口数和是否工作日2个因素.通过10折交叉验证表明，RBF神经网络回归模型的拟合效果比多元线性回归模型更好.

Taxi on Service Trip Characteristics Based on Multi-source Data Fusion: A Case of Yueyang

In order to explore the characteristics of taxi on service, those are fused that resident trip survey data, urban land use data and weather data, basis on the large data of taxi GPS trajectory. A passenger taxi travel volume regression model is constructed. It is concluded that there is a strong correlation between the passenger travel volume and the number of posts, weather conditions, time period, area of the district. Regression model and RBF neural network is constructed based on the above four factors on the increase in the district of the resident population and whether weekdays. Through 10 fold cross validation indicate that the fitting effect of RBF neural network model is better than multivariate linear regression model.