Affiliation:
1. University of Technology and Applied Sciences, Oman
2. University of Naples, Italy
3. Bennett University, India
Abstract
Forecasting future trends in tourism growth is imperative for sustainability planning, yet highly complex due to the sector's multifaceted nature. This study leverages machine learning techniques to develop an integrated model predicting foreign tourist arrivals to India. Utilizing 2000-2022 data encompassing tourist statistics alongside relevant socioeconomic indicators, advanced algorithms like XGBoost uncover key drivers and relationships to generate strategic long-range forecasts. The multilayered analysis reveals tourism infrastructure investments strongly stimulate arrivals, underscoring policy priorities. However, skills training expenditures exhibit a more nuanced linkage, indicating localized needs. Beyond forecasting accuracy, the research makes significant methodological contributions regarding multivariate input features and model robustness for tourism ecosystems. It advocates systems thinking-based approaches over reductionist modeling of isolated past arrivals, given tourism's interdependence with broader socioeconomics.