计及复杂影响因素的区域大数据智能预测算法设计

Design of intelligent prediction algorithm for large area data considering complex influencing factors

针对传统客流量分析预测方法存在运行效率低且难以对数据隐藏特性进行深入分析的问题,文中基于改进长短时神经网络模型提出了一种区域性的游客流量预测算法。该算法对传统长短时神经网络的神经元进行了简化,有效提高了收敛速度。同时采用自编码器对输入数据特征进行提取,从而得到了具有更强关联性的数据。在模型输出部分,通过引入多头注意力机制对输出数据进行权重排序,以获得更准确的预测结果。实验测试结果表明,所提算法的RMSE值为15.81,运行时间仅为30 min,且其客流量预测误差在所有对比算法中最小,可以为区域游客流量预测提供数据支撑。

In view of the shortcomings of traditional passenger flow analysis and prediction methods,such as low operating efficiency and difficult to deeply analyze the data hiding characteristics,this paper proposes a regional tourist flow prediction algorithm based on the improved Long Short Time Neural network model.The algorithm simplifies the neurons of the traditional Long Short Time Neural network,and effectively improves the convergence speed.At the same time,the self encoder is used to extract the features of the input data,so that the data with stronger relevance is obtained.In the output part of the model,the multi head attention mechanism is introduced to rank the output data by weight,and a more accurate prediction result is obtained.The test results show that the RMSE of the algorithm in this paper is 15.81 and the running time is only 30 minutes.The prediction error of passenger flow is the smallest among all the comparison algorithms,which can provide data support for the prediction of regional tourism passenger flow.