Research On Network Operation Situation Analysis And Prediction Technology Based On Fuzzy Association Rule Mining

With the rapid development of information technology in the 21 st century,great changes have taken place in many industries after the emergence of confidence technology.People's Daily life,study and work begin to rely more and more on the existence of the network.In order to reduce the loss caused by the deterioration of network operation condition and even avoid its loss,network operators hope to know the operation quality of the network at any time so as to take relevant measures in the first time.Network operation state is a comprehensive and objective evaluation result of network operation state.Network operation situation is a comprehensive reflection of whether the whole network is in normal operation,whether each device in the network is in normal operation,and whether network users can use network services normally.The ultimate goal of network operation situation analysis is to obtain the operation situation results and changes of the network in the running time so that the network administrator can intuitively understand the operation status and stability of the complex network.Correlation analysis can mine the potential correlation between nodes,reflect the interactional relationship within the network,evaluate and predict the network operation state from the whole network,and use fuzzy technology to solve the uncertainty problem in the network operation analysis and evaluation.The research features of this paper are the combination of fuzzy theory technology and association rule mining to form a fuzzy association rule mining method to analyze and predict the network operation situation.The main work and innovation are as follows:1.Built the analysis model of network operation situation awareness,established the index system of network operation situation assessment,and designed the specific process of data preprocessing.2.Fuzzy clustering was used to propose network operation state characteristics,and fuzzy processing was carried out on characteristic data,and a fuzzy database describing node operation state was established.An improved adaptive fuzzy clustering algorithm,AFCM,is proposed in this paper.3.A new fuzzy association rules mining algorithm FPLI is proposed.The algorithm adopts the data structure of linear linked list,introduces the interest-based constraint and adopts the working mode of parallel mining.Through simulation verification and comparative analysis,the algorithm has strong advantages over traditional mining algorithms in terms of time complexity,mining efficiency and interest constraint.In this paper,the improved algorithm is applied to the evaluation and analysis of the network operation situation,and the results can objectively reflect the network operation situation.4.The situation prediction system based on fuzzy association rules is established,and the genetic algorithm is used to optimize the fuzzy prediction system to improve the prediction accuracy and simplification.Through simulation verification and comparative analysis,the effectiveness of this method in network operation situation prediction is better than the traditional time series prediction method.This paper studies the network operation situation analysis,proposes a series of new algorithms for network operation situation assessment and prediction,and performs simulation verification.The results reach the expected goal.