基于疫情影响构建的旅游需求预测模型  

Forecasting Model of Tourism Demand Based on COVID-19 Impacts Construction

作　　者：杨京津 

机构地区：[1]兰州交通大学交通与运输学院,甘肃兰州730070

出　　处：《工业控制计算机》2023年第7期87-88,共2页Industrial Control Computer

摘　　要：鉴于旅游业对环境波动的高度敏感性,提出了考虑疫情影响的旅游需求预测框架。首先,收集目标城市的过夜游客人数作为旅游需求数据。其次,基于数据在疫情发生前后的数据波动,提出疫情影响构建和核心信息提取方法,该方法基于BHT-ARIMA算法,能够有效提取数据中潜藏的疫情影响特征,并有效剔除了数据冗余。最后,基于CNN-GRU模型进行模型训练和预测,得到最终的预测结果。为了验证提出框架的有效性,在旅游城市三亚的过夜游客人数数据集上进行了实验。实证结果表明,与其他基准模型相比,所提出的框架具有显著的预测性能,预测误差低至0.0991。Given the high sensitivity of tourism to environmental fluctuations,this paper proposes a tourism demand forecasting framework that takes into account the quantification of the impact of COVID-19.First,the number of overnight tourists in the target city is collected as tourism demand data.Secondly,based on the data fluctuations before and after the outbreak of the epidemic,a method of epidemic impact quantification and core information extraction is proposed.Based on the BHT-ARIMA algorithm,this method could effectively extract the hidden epidemic impact characteristics in the data and effectively eliminate the data redundancy.Finally,model training and prediction are carried out based on CNN-GRU model to obtain the final prediction results.

关 键 词：旅游需求预测 核心信息提取 BHT-ARIMA CNN-GRU 

分 类 号：F224[经济管理—国民经济] F592.7