BDT: An Ada Boost Classifier Ensemble with Decision Tree for Traffic Network Prediction
N. Deeban, P. Shyamala Bharathi · 2022
The identification of network traffic is now one of the most active study areas in the areas of network management and network security. Machine learning is a key technology that is utilized during the identification of network traffic research. The first step in analyzing and determining the various kinds of applications that are moving across a network is to classify the traffic flowing through the network. Internet service providers and other operators of networks can better manage the overall functioning of their networks by utilizing this method. Traditional techniques for classifying internet traffic include things like port-based, pay-load-based, and machine-learning-based methods. These methods are only a few of the many available. The Machine Learning (ML) methodology, which is employed by a great number of researchers and has gotten extremely effective accuracy results, is now the method that is the most often used. In this study, we have concentrated on evaluating network data traffic for 5G Network utilizing an ensemble approach known as AdaBoost ensemble with Decision Tree (BDT). To decrease training errors, boosting is a sort of ensemble learning that integrates a set of weak learners into a single powerful learner. AdaBoost works well with decision trees with only one level, hence these are the methods most often used. On the basis of the test data, the classification accuracy is an average of 98.2 percent, significantly surpassing other approaches.