Research on Stability of SAAS Application Software Based on Dynamic Fault Prediction of Internet of Things

With the maturity of the deployment mode of Internet of things services, the application fields are more and more extensive. People have higher requirements on the availability of network platform services, hoping to obtain uninterrupted network platform services. SAAS service mode can upload all kinds of data needed by customers to SAAS platform for storage. Many of these data are business sensitive data. Once a data security accident occurs, the impact will be very bad. This paper proposes a network data security optimization algorithm based on compactness. Firstly, in order to consider the uneven distribution of training samples and the influence of noise samples on classification accuracy, a method based on the closeness degree between samples is proposed to calculate membership degree. Then, the fuzzy entropy value of the feature is used to determine the weight of each sample feature, and the weighted Euclidean distance based on the feature weight is used to determine the nearest neighbor of the sample to be classified, so as to better reflect the difference of each sample feature. Finally, the classification of the samples to be classified is determined according to the membership degree of each category. Through experimental comparison test, the prediction accuracy of the research method in this paper is significantly higher than that of other comparison methods, with higher practical effect, which can provide a stable operation guarantee for the SAAS platform.