Optimization and Prediction Algorithm of Network Security Situation Based on Big Data Analysis

In response to the time variability and nonlinearity issues in predicting the current network security situation, this article introduces big data analysis technology and combines it with the Cloud Computing Online Extreme Learning Machine (CC-OELM) model to optimize the prediction algorithm of network security situation, improve the accuracy and stability of prediction. Firstly, this article analyzes existing network security prediction models and identifies the limitations of traditional methods in the big data environment. Then, the CC-OELM model is designed and optimized, dividing the network security situation data into multiple sub training sets. Finally, experiments conducted on large-scale samples have shown that the optimized model has significant advantages in dealing with nonlinear and dynamically changing network security issues. Through comparative experiments on multiple models, when the number of samples is 4000, the expansion rate is 0.271. This model can maintain good scalability in large-scale data processing, providing theoretical support and practical basis for active defense of network security.