Forecast Research Of Inbound Tourism Demand Based On Artificial Neural Network

Artificial neural network is an important branch of Deep learning.Due to its wide adaptability and learning ability,it has been widely used and developed in the prediction of nonlinear systems.The related research of tourism demand prediction model is one of the important research topics in tourism studies.Inbound tourism is a great importance in balancing international balance,increasing foreign exchange income and enhancing the international influence of China's tourism culture.Compared with the traditional predictive analysis model,the artificial neural network model is better at dealing with multidimensional nonlinear problems;in addition,the data related to tourism demand prediction is very comprehensive,and the statistical time is short.As tourism demand is affected by various complex factors,the forecast analysis model based on artificial neural network is more superior.Firstly,the paper analyzes the research status of domestic and foreign scholars in the field of artificial neural network application and tourism demand forecasting,and comprehensively analyzes the application status of artificial neural network in the field of tourism demand forecasting.Secondly,it analyzes the basic principle and application of traditional forecasting method,time series method.Regression prediction methods and Grey prediction methods are mainly applied to the study of structured data and linear relations.For the study of nonlinear problems,such as artificial neural networks,machine learning methods are more suitable.Therefore,artificial neural networks are used in the forecast analysis of inbound tourism demand.The method is used to establish the predictive model;then the influencing factor of the inbound tourism is selected to construct the feature vector.Since the tourism related data statistics are not complete and accurate,and the starting time is late,the constructed feature vector has certain difficulty in quantification.It includes ten factors such as the number of inbound tourists,per capita GDP,total import and export of goods,final consumption rate of residents and population.On this basis,BP neural network,RBF neural network,SVM neural network and GA-SVM neural network structure are constructed and artificial neural network prediction model is constructed.The principle of four models and the process of network design and implementation are analyzed in detail.In the empirical study of the forecast of inbound tourism demand,the data of artificial neural network learning mainly comes from the time series-based data of China's major international tourist source countries from 2007 to 2016 in China Statistical Yearbook.The mean square error MSE and the determination coefficient were selected as the evaluation indexes of the model performance.The learning training results of four artificial neural network models,BP neural network,RBF neural network,SVM neural network and GA-SVM neural network,show that the GA-SVM neural network algorithm improved by genetic algorithm is the best.The MSE indicator in the test set is 0.010775,which is 0.96748,indicating that the accuracy of the training process is very high.The improved neural network requires less learning samples and exhibits excellent learning and prediction ability when the sample size is small.Therefore,the GA-SVM neural network model is used as a predictive model for the inbound tourism demand trend from 2017 to 2019.After learning the sample data,the trained GA-SVM neural network algorithm is used to predict the development trend in the next three years.According to the prediction results,the development of the demand of China's five major source countries for tourism in China is discussed and related.Finally,the paper analyzes the full text and points out the inadequacies in the article and the follow-up research prospects of the paper.Artificial neural network is a good predictive model,which can be applied to the problem of inbound tourism demand forecasting.The forecasting research on inbound tourism demand is conducive to enriching and developing tourism related theories,in tourism policy formulation and tourism product development.There is a certain reference.