Unmanned aerial vehicles, or UAVs, have entirely transformed various industries, including agriculture, military, logistics, medical, and surveillance. With increased usage comes increased susceptibility to sophisticated attacks, which pose significant risks to operational efficiency and data integrity. UAVs are being used extensively in various applications, which has increased the need for strong security measures because these devices are vulnerable to sophisticated cyberattacks like replay, denial-of-service, and fake data injection. The detection capabilities of the existing intrusion detection systems (IDS) are limited since they frequently concentrate on either physical or cyber data. Developing a unique IDS that combines physical and cyber data to enhance detection using machine learning is possible. We compared the performances of several models and used the Python library Lazypredict featuring the LazyClassifier and LazyRegressor to examine various models at once, which reduces time and helps to choose the best model for developing the system. Extensive analyses of separate and hybrid cyber-physical datasets demonstrated that models trained on integrated data outperformed those based on cyber or physical data alone, particularly when faced with novel or unexpected attacks. The integration of numerous data sources improved the performance of IDS and allowed for a more accurate overview of UAV operations.