Abstract
The tourism forecasting industry is facing challenges in improving its performance. However, this presents an opportunity for research and development, especially for developing countries looking to improve their tourism sector. The objective of a systematic review was identifying the newest and most used models for forecasting tourism demand. Using the PRISMA method and analyzing 40 articles, we concluded that hybrid models are the most effective for tourism demand forecasting. Combining a statistical model with a machine learning-based model is particularly efficient.