Volume 11 - Volume 11
Detecting and Analysing Network Logs Using Machine Learning Techniques
Abstract
A Network rhetorical investigation needs the capturing, recording and analysing of network proof
and therefore the audit trails. The end result of such investigations could deliver security audit,
security data to harden a system or proof for legal functions. One of the most common method of
attack involve sending large amount of request to sites or server and server will be unable to handle
and sites will be offline for many days. The aim is at determining the verity by analysing the network
activity logs. Some machine learning algorithms are deployed to find various attack like DoS UR2,
R2L and probe type. The mooted machine learning algorithms are SVM, Random Forest, Decision
Tree, Logistic Regression, Naïve Bayes and K-nearest neighbour. Multiple network logs are collected
and analysed using Wireshark. Log analysis is then availing to find the quandary in period of time
and fine-tune it before it engenders ruin. A mechanism designed to aggregate and analyses these logs
to have a clear overview of what’s transpiring across the network to determine various attacks. And
finally, the effectiveness and accuracy of various machine learning algorithms on log datasets having
particular features are evaluated. Thus, indemnity from attackers could be enhanced by providing
dependability, solidity and surety.
Paper Details
PaperID: 1937
Author's Name: Akhila Anilkumar, Alona Shibu, Meera Anna Varghese, Priya P Sajan and A.L. Sreedeep
Volume: Volume 11
Issues: Volume 11
Keywords: Network Logs, Wireshark, Log Analysis, Machine Learning Techniques.
Year: 2021
Month: May
Pages: 271-286