Research on filling algorithm of incomplete data in north interface of optical fiber network

Abstract

Northbound interface is the interface for manufacturers or operators to access and manage the network. Through it, the superior network management can obtain a large number of data such as configuration performance and operation and maintenance of optical network. For various reasons, there are usually missing values in real data sets. In order to improve the accuracy of filling missing values. In this paper, from the perspective of data mining, we use a variety of processing methods to fill the missing value of the north interface. In this paper, by analyzing the principle of KNN algorithm, decision tree algorithm, random forest algorithm and extreme random tree algorithm, the four algorithms are simulated in the missing value processing of northbound interface data respectively. The performance differences of various algorithms are compared and analyzed, and their advantages and disadvantages are compared. Finally, the most suitable algorithm for missing value processing of northbound interface data is found. Based on the simulation and experimental results, it is concluded that the extreme random tree algorithm has better filling effect in dealing with the missing values of the north interface.