scholarly journals Prospects for the Application of Reinforcement Learning to Network Traffic Classification Tasks

2021 ◽  
Vol 2096 (1) ◽  
pp. 012175
Author(s):  
G D Asyaev
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Reinforcement Learning ◽  
Network Traffic ◽  
Classification Task ◽  
Traffic Classification ◽  
Data Set ◽  
Basic Principles ◽  
Network Traffic Classification ◽  
Classification Tasks

Abstract The basic principles and methods of reinforcement learning are reviewed. The problems and approaches for applying a model based on reinforcement learning in the framework of attack prevention are described. The model is built and the hyperparameters of machine learning for the task of classifying network traffic are selected, and its performance on the test data set is evaluated by such quality metrics as accuracy and completeness. The dataset used to implement an agent for selecting the optimal defense strategy for a particular attack has been finalized. Developed an algorithm for using a reinforcement learning neural network for the traffic classification task. A table of rules and rewards for the problem is generated. An agent has been developed and trained to interact with the system. We describe the application of reinforcement learning to the traffic classification task.

scholarly journals MODC: A Pareto-Optimal Optimization Approach for Network Traffic Classification Based on the Divide and Conquer Strategy

Information ◽  
2018 ◽  
Vol 9 (9) ◽  
pp. 233 ◽  
Author(s):  
Zuleika Nascimento ◽  
Djamel Sadok
Keyword(s):  
Machine Learning ◽  
Network Traffic ◽  
Machine Learning Algorithms ◽  
Divide And Conquer ◽  
Pareto Optimal ◽  
Optimization Approach ◽  
Traffic Classification ◽  
Multi Objective ◽  
Network Traffic Classification ◽  
Changes Over Time

Network traffic classification aims to identify categories of traffic or applications of network packets or flows. It is an area that continues to gain attention by researchers due to the necessity of understanding the composition of network traffics, which changes over time, to ensure the network Quality of Service (QoS). Among the different methods of network traffic classification, the payload-based one (DPI) is the most accurate, but presents some drawbacks, such as the inability of classifying encrypted data, the concerns regarding the users’ privacy, the high computational costs, and ambiguity when multiple signatures might match. For that reason, machine learning methods have been proposed to overcome these issues. This work proposes a Multi-Objective Divide and Conquer (MODC) model for network traffic classification, by combining, into a hybrid model, supervised and unsupervised machine learning algorithms, based on the divide and conquer strategy. Additionally, it is a flexible model since it allows network administrators to choose between a set of parameters (pareto-optimal solutions), led by a multi-objective optimization process, by prioritizing flow or byte accuracies. Our method achieved 94.14% of average flow accuracy for the analyzed dataset, outperforming the six DPI-based tools investigated, including two commercial ones, and other machine learning-based methods.

scholarly journals Robot Communication: Network Traffic Classification Based on Deep Neural Network

2021 ◽  
Vol 15 ◽  
Author(s):  
Mengmeng Ge ◽  
Xiangzhan Yu ◽  
Likun Liu
Keyword(s):  
Neural Network ◽  
Network Traffic ◽  
Short Term Memory ◽  
Traffic Classification ◽  
Classification Framework ◽  
Network Traffic Classification ◽  
Scalar Input ◽  
Memory Network ◽  
Encrypted Traffic ◽  
Robot Communication

With the rapid popularization of robots, the risks brought by robot communication have also attracted the attention of researchers. Because current traffic classification methods based on plaintext cannot classify encrypted traffic, other methods based on statistical analysis require manual extraction of features. This paper proposes (i) a traffic classification framework based on a capsule neural network. This method has a multilayer neural network that can automatically learn the characteristics of the data stream. It uses capsule vectors instead of a single scalar input to effectively classify encrypted network traffic. (ii) For different network structures, a classification network structure combining convolution neural network and long short-term memory network is proposed. This structure has the characteristics of learning network traffic time and space characteristics. Experimental results show that the network model can classify encrypted traffic and does not require manual feature extraction. And on the basis of the previous tool, the recognition accuracy rate has increased by 8%

scholarly journals Towards the Deployment of Machine Learning Solutions in Network Traffic Classification: A Systematic Survey

2019 ◽  
Vol 21 (2) ◽  
pp. 1988-2014 ◽  
Author(s):  
Fannia Pacheco ◽  
Ernesto Exposito ◽  
Mathieu Gineste ◽  
Cedric Baudoin ◽  
Jose Aguilar
Keyword(s):  
Machine Learning ◽  
Network Traffic ◽  
Traffic Classification ◽  
Systematic Survey ◽  
Network Traffic Classification

scholarly journals Voting Classification Method for Network Traffic Prediction

2020 ◽  
Vol 9 (7) ◽  
pp. 510-515
Keyword(s):  
Network Traffic ◽  
Second Phase ◽  
Traffic Classification ◽  
Data Set ◽  
Prediction Analysis ◽  
Proposed Model ◽  
Network Traffic Classification ◽  
Target Set ◽  
The Relationship ◽  
Complex Dataset

The prediction analysis is the approach of data mining which is applied to predict future possibilities based on the current information. The network traffic classification is the major issue of the prediction analysis due to complex dataset. The network traffic techniques have three steps, which are preprocessing, feature extraction and classification. In the phase of pre-processing data set is collected which is processed to removed missing and redundant values. In the second phase, the relationship between attribute and target set is established. In the last phase, the technique of classification is applied for the classification. This research study has been influenced by the different intrusion threats on internet and the ways to detect them. In this research, we have studied and analyzed the famous network traffic data -NSL KDD dataset and its various features. The proposed model is a hybrid of Logistic Regression and Knearest neighbor classifier combined using voting classifier, which aims at classifying the data into malicious and nonmalicious with more accuracy than existing methods.

Sign in / Sign up

Export Citation Format

Share Document