Economic implications of deep machine learning for tourism time
Economic implications of deep machine learning for tourism time

Summary/Abstract: Research background: Predicting the flow of inbound tourism accurately has always posed a significant challenge for all parties involved in the industry. The complex nature of the tourism product, which is directly and indirectly affected by various risks, disasters, and crises, further highlights its susceptibility to disruptions and fluctuations. As a result, there has been a growing interest in forecasting inbound tourism flows using contemporary data science methods and artificial neural network (ANN) methods. This paper, therefore, seeks to explore the AI forecasting techniques by employing a deep machine learning (DML) approach and comparing various Python libraries for time series forecasting within a Jupyter Notebook computing environment. The data from domestic and international tourism stays registered in Bulgaria for the period 2002 to 2023 has been utilized to construct an advanced deep neural network using multiple Python libraries. Purpose of the article: The purpose of the current paper is to establish which time series forecasting model - Exponential Smoothing, TBATS, Auto ARIMA, Theta or LSTM has better accuracy estimation and may be applied for similar tasks in the future by for research and practical economic purposes. Methods: The applied methodology was based on the classical scientific method. As to the main findings they could help in daily operation planning, managing and relocating resources in tourism on micro and macro level as well as with insights on the drawbacks and limitations on this research area related to analysis and novelty implications. Findings & Value added: What is more, the results obtained reaffirmed that ANN can be applied for an accurate forecasting, especially in the case of Bulgaria, where such models have not been applied yet neither by the tourism related academics nor by the business and the policymakers.

• Details
• Contents