旅游需求量预测算法研究

Research on tourism demand prediction algorithm

空置的酒店房间、未售出的活动门票和未被消费的商品等代表了不必要的成本和未实现收入,政府需要准确的对旅游需求量加以预测,以便在基础设施开发和住宿场地规划等问题上做出明智的决策,因此准确预测旅游需求量变得至关重要。随着人工智能的迅速发展,神经网络和SVR等人工智能模型在旅游需求量预测中得到了成功的应用。本文将对SVR,BP神经网络,LSTM三种旅游需求量预测模型的训练过程、应用和特点进行研究,对其中实际难点进行了分析,总结和展望了旅游需求量预测模型在实际生活中的应用。

Empty hotel rooms,unsold event tickets and unconsumed goods represent unnecessary costs and unrealized revenues,and accurate forecasts of travel demand are essential for governments to make informed decisions on issues such as infrastructure development and accommodation planning.With the rapid development of artificial intelligence,artificial intelligence models such as neural network and SVR have been successfully applied in the prediction of tourism demand.In this paper,the training process,application and characteristics of SVR,BP neural network and LSTM tourism demand prediction models are studied,and the practical difficulties are analyzed,and the application of tourism demand prediction models in real life is summarized and prospected.