基于双层K近邻算法的短时交通流预测

双层K近邻算法在K近邻算法的基础上,增加了模式匹配步骤,从而提高了K近邻算法的预测精度.鉴于此,利用双层K近邻算法,对北京市微波检测器数据进行分析,进而标定算法的最优参数.同时定义了预测算法的滞后性,并将双层K近邻算法与自适应预测算法的滞后性进行了对比,从预测精度及滞后性两方面验证了双层K近邻算法的适用性.

Short-Term Traffic Flow Forecasting Based on Two-Tier K Nearest Neighbor Algorithm

Based on traditional K nearest neighbor algorithm, Two-Tier K nearest neighbor algorithm adds the pattern matching step to improve the prediction accuracy of K nearest neighbor algorithm. Given this, the paper uses Two-Tier K nearest neighbor algorithm to analyze microwave detector data of Beijing and calibrate the algorithm parameters. Meanwhile, it defines the lag of prediction algorithm, and contra- sts the prediction lag of Two-Tier K nearest neighbor algorithm and self-adaptive filtering method to veri- fy the applicability of Two-Tier K nearest neighbor algorithm from aspects of prediction accuracy and lag.