基于深度学习的网约车供需缺口短时预测研究

城市不同区域网约车供需缺口预测可为车辆调度策略提供支持，从而提高车辆运行效率和乘客服务水平.为实现网约车供需缺口短时预测，提出一种基于时空数据挖掘的深度学习预测模型(Spatio-Temporal Deep Learning Model, S-TDL).该模型由时空变量模型、空间属性变量模型和环境变量模型 3个子模型融合而成，可捕捉时空关联性、区域差异性和环境变化对供需缺口的影响.同时，提出特征聚类-最大信息系数两阶段特征选择方法，筛选与供需缺口相关性强的特征变量，提高训练效率，减少过拟合.滴滴出行实例分析证明，特征选择后的 STDL模型预测精度显著优于BP神经网络、长短期记忆网络和卷积神经网络.

Short-term Forecasting of Supply-demand Gap under Online Car-hailing Services Based on Deep Learning

The results of supply- demand gap prediction for online car- hailing services in different areas can provide support for online car-hailing scheduling system, thereby improving efficiency and service levels. In order to realize the short-term forecast of supply-demand gap for online car-hailing services, this paper proposes a novel spatio- temporal deep learning model (S- TDL). The model is composed of three sub- models: spatiotemporal variable model, spatial attribute variable model and environment variable model. It can capture the impact of spatio- temporal correlation, regional difference and environmental change on supply- demand gap. Moreover, a feature selection method named feature clustering-maximum information coefficient two-stage feature selection is proposed to screen out the important features which are strongly correlated with the supply- demand gap, improve training efficiency. The experimental results show that the S-TDL model after feature selection achieves the better performance than the existing methods.