Deep Machine Learning for Time Series Inbound Tourism Forecasting
Deep Machine Learning for Time Series Inbound Tourism Forecasting

Author(s): Ivanka Vasenska
Subject(s): Social Sciences
Published by: Udruženje ekonomista i menadžera Balkana
Keywords: Time series; Deep machine learning; Artificial intelligence; Bulgaria inbound tourism forecast
Summary/Abstract: Accurate inbound tourism flow forecasting has been a challenge for all stakeholders related to the sector. The multidisciplinary character of the tourism product which has been directly and indirectly influenced by all types of risks, cataclysms and crises further exposed its intangible nature to shocks and flows disruption. Thus, forecasting inbound tourism flows with advanced data science and AI (artificial intelligence) methods has been gaining momentum, which the COVID-19 pandemic boosted. Therefore, this paper aims to examine the relevant AI forecasting methods by applying a deep machine learning technique comparing different Python time series forecasting libraries via a Jupyter Notebook computer environment. Bulgaria’s inbound tourism data has been used to develop an advanced deep neural network with the DARTS Python library and compare its accuracy with other Python library models.

• Details
• Contents