A Smart System for Detecting Behavioural Botnet Attacks using Random Forest Classifier with Principal Component Analysis

Taylor, O. E.; Ezekiel, P. S.

doi:10.24018/ejai.2022.1.2.4

Cited by 3 publications

(2 citation statements)

References 6 publications

(6 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The accuracy rates for the several categories are as follows: normal at 0.79%, DoS at 0.94%, R2L at 0.88%, Probe at 0.94%, and U2R at 0.99%. In the study conducted by [27], a sophisticated method was proposed for the identification of behavioural bootnet attacks. This system leveraged on the Random Forest Classifier and Principal Component Analysis (PCA) techniques.…”

Section: Literature Reviewmentioning

confidence: 99%

Untitled

2024

INTERNATIONAL JOURNAL OF ADVANCED RESEARCH IN COMPUTER AND COMM

View full text Add to dashboard Cite

The utilization of machine learning has significant importance in the identification and prevention of distributed denial of service (DDoS) attacks. Through the examination of network traffic patterns, machine learning algorithms possess the capability to detect anomalous activities that serve as indicators of a Distributed Denial of Service (DDoS) assault in a timely manner. This paper presents a novel hybrid approach for the detection of distributed denialof-service (DDoS) attacks in network logs, leveraging the strengths of both Random Forest (RF) for feature extraction and Recurrent Neural Network (RNN) for classification. The proposed framework harnesses the discriminative power of RF in identifying salient features from the raw network log data, which are subsequently utilized as input for the RNN classifier. The Random Forest algorithm was employed to extract a comprehensive set of discriminative features from the network log data, enabling the model to capture intricate patterns indicative of DDoS attacks. These features were then employed as input to the RNN classifier, facilitating the utilization of sequential dependencies and temporal patterns within the log data. The hybrid model achieved exceptional performance, with an accuracy of 99.99%. Furthermore, the true positive rate was recorded at an impressive 99.99%, demonstrating the model's proficiency in correctly identifying actual instances of DDoS attacks. The false positive rate was exceptionally low, at 0.0001%, underscoring the model's robustness in minimizing misclassifications. This study represents a significant advancement in the field of DDoS attack detection, offering a powerful and accurate solution that effectively combines the strengths of Random Forest for feature extraction and RNN for classification. The hybrid model's outstanding performance metrics affirm its potential for deployment in real-world network security environments, providing a robust defense against DDoS attacks.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Untitled

2024

INTERNATIONAL JOURNAL OF ADVANCED RESEARCH IN COMPUTER AND COMM

View full text Add to dashboard Cite

show abstract

“…In [14], a single objective genetic algorithm is employed to explore the large space of candidate attributes to identify the optimal subset that could boost the efectiveness of the GAs-based botnet detection method. Te authors of [15] describe the C4.5 algorithm-based botnet fow classifcation system.…”

Section: Related Workmentioning

confidence: 99%

Smart Approach for Botnet Detection Based on Network Traffic Analysis

Obeidat

Yaqbeh

2022

Journal of Electrical and Computer Engineering

View full text Add to dashboard Cite

Today, botnets are the most common threat on the Internet and are used as the main attack vector against individuals and businesses. Cybercriminals have exploited botnets for many illegal activities, including click fraud, DDOS attacks, and spam production. In this article, we suggest a method for identifying the behavior of data traffic using machine learning classifiers including genetic algorithm to detect botnet activities. By categorizing behavior based on time slots, we investigate the viability of detecting botnet behavior without seeing a whole network data flow. We also evaluate the efficacy of two well-known classification methods with reference to this data. We demonstrate experimentally, using existing datasets, that it is possible to detect botnet activities with high precision.

show abstract

Timely detection of DDoS attacks in IoT with dimensionality reduction

Kumari,

Jain

2024

Cluster Comput

View full text Add to dashboard Cite

A Smart System for Detecting Behavioural Botnet Attacks using Random Forest Classifier with Principal Component Analysis

Cited by 3 publications

References 6 publications

Untitled

Untitled

Smart Approach for Botnet Detection Based on Network Traffic Analysis

Timely detection of DDoS attacks in IoT with dimensionality reduction

Contact Info

Product

Resources

About