关键词: 丢失数据, 补齐, 最小二乘支持向量机, 时间序列, 交通流, 智能运输系统

Abstract: Real-time and accurate traffic flow data is essential for Intelligent Transportation Systems research.Traffic control and traffic guidance are directly affected by the quality of input data.The detected data is often incomplete and may cause out of order.The model for filling time series data of traffic flow based on LS-SVM is proposed in this paper，missing data can be filled by using traffic flow historical data.The simulation results show that LS-SVM have better generalization ability and strong robustness.

Key words: missing data, filling, Least Squares Support Vector Machines（LS-SVM）, times series, traffic flow, Intelligent Transportation Systems（ITS）