Neural network-based prediction of domestic tourists' length of stay in Italy

This study proposes a predictive approach to estimate the duration of domestic tourists' stays in Italy, utilizing microdata from the 2019 Travel and Holiday Survey provided by the Italian National Institute of Statistics (ISTAT). The main objective is to develop a predictive model using neural networks and deep learning methodology, leveraging the most relevant predictive variables identified in the scientific literature. Deep learning methodology offers significant advantages in detecting and interpreting complex patterns in data, overcoming the limitations of traditional models. The aim is to achieve an accurate and reliable predictive model, providing valuable insights for the tourism sector and tourism development policies. This study contributes to the understanding and prediction of the duration of domestic tourists' stays in Italy, fully exploiting the benefits of deep learning methodology. The obtained results will have a significant impact on the tourism sector, enabling the optimization of hospitality and tourism promotion strategies in the country. By combining the microdata provided by ISTAT with the deep learning approach, complex relationships between the predictive variables and the duration of the tourist stay can be identified providing a robust foundation for informed decisions in the field of tourism.