A machine learning based cyber security threat prediction algorithm for tourism hotels

In the travel and hospitality sector, digitalisation has brought network security threats. Particularly with big data and high-dimensional characteristics, traditional network security techniques find it difficult to control dynamic and complicated security threats. Popular research subjects based on performance are machine learning-based threat prediction methods and integrated learning approaches. This paper presents XG-CatSec, a machine learning (XGBoost and Catboost fusion) model to increase tourist and hospitality cybersecurity threat prediction accuracy and robustness. While CatBoost simplifies data preparation and optimises category feature processing, XGBoost increases model accuracy utilising gradient boosting trees. Combining these technologies in XG-CatSec raises the threat identification for hotel cybersecurity. XG-CatSec beats SVM and random forest in accuracy, precision, and recall on the NSL-KDD data. This report motivates further research by implying a special cybersecurity threat prediction solution for tourism and hospitality.