The study of tourism ecological security is a core problem in the research of sustainable tourism development. It has important theoretical significance and practical value for coordinated development of island tourism economy and the ecological environment. It can be used to scientifically predict island destination development trends of tourism ecological security. Tourism destination can be regarded as an organism with a complex ecosystem. To our knowledge, once the function is disordered, the destination must be considered threatened. The purpose of this paper was to construct a tourism ecological safety index system based on the five subsystems, including "Carrying Capability", "Supporting Capability", "Attraction Capability", "Evolution Capability", and "Developing Capability" (Known as the CSAED model). Based on the sustainable development perspective, and the comprehensive, dynamic principle; this paper uses Mount Putuo Island in Zhejiang Province as an example, and constructs a tourism ecological safety index system based on the subsystems of carrying capacity, attraction capability, evolution capability, and development capability (CSAED model). The paper used the Grey system GM (1, 1) model and the radial basis function (RBF) neural network model to forecast tourism ecological safety in Mount Putuo Island. The results showed that:(1) both the Pearson correlation coefficient for the RBF neural network and the root mean squared error were better than the Grey System GM (1, 1) model. They also exhibited a better linear fit and a higher precision of prediction. This paper used Grey Relational Analysis to select the main driving factors; and used the results of linear and nonlinear analysis to build equations for trend extrapolation. In addition, based on the results of principal component analysis, the RBF neural network model appeared to provide a new research area for tourism destination ecological security. One of the key issues was that the tourism ecological security situation of Mount Putuo Island became better, because the index of the RBF model from 2005 to 2014 predicted results from 0.3568 to 0.6475. It appeared that the security level increased the sensitivity level, critical level, and the general level. Additionally, the index of the RBF model from 2015 to 2020 predicted results were from 0.7010 to 0.8135, the security level increased from the relatively safe grade to very safe grade. However, it is well known that the ecological system on the island will be influenced by several factors, including natural, social, and economic, among others. In terms of Mount Putuo Island, during the period of the forecast, it may be affected by typhoons, which may affect the vulnerability of the tourism industry. As such, it may lead to deviation from forecasted results. In short, it is suggested that the perspective of natural ecosystems be considered in future research, which would help to construct a better tourism ecological security index system through better mathematical algorithms to enhance the progress of a RBF neural network model. Consequently, the results could provide critical suggestions to scientifically protect tourism ecological security in Mount Putuo Island.