A Combined Prediction Method of Industrial Internet Security Situation Based on Time Series

Abstract

Attacks on industrial control systems have different types and various intensities. Predicting the development trend of security state for updating defense strategy in time is important. Every single prediction model has its different emphases, and the accuracy of single prediction model may be reduced when the system is attacked. Combined prediction model integrates the prediction results of multiple single prediction models to improve the overall prediction accuracy. The weight of the combined model is determined by the mean square error of every single model. It is taken a logarithmic function of the mean square error of each single model to enlarge the difference between the superior and inferior models. Then, the simple weight function is taken to determine the weight of each model based on the logarithm result of each model.This approach makes greater use of accurate model information. Through the comparative analysis of the model, the error of the combined prediction method is obviously reduced. Through the comparison and analysis of the weighted method, the error of the weighted method of the combined prediction model proposed in this paper is the minimum. The combined prediction method can provide more accurate defense opinions for network security administrators.