DETECTION OF CYBER ATTACK IN NETWORK USING MACHINE LEARNING TECHNIQUES

Abstract

In contrast to the past, advancements in personal 
computer and communication technologies have 
brought about significant changes. Although using 
new technology gives individuals, organisations, and 
governments enormous benefits, some people are 
messed up against them. For instance, the 
safeguarding of important data, the safety of 
information transfer channels, the availability of 
information, and so on. In light of these problems, 
digital oppression motivated by fear is one of the 
biggest problems we face today. Digital dread, which 
caused a lot of problems for individuals and 
organisations, has reached a point where it might 
compromise national and open security due to many 
groups, including the criminal underworld, 
professionals, and digital activists. As a result, 
Intrusion Detection Systems (IDS) were developed to 
keep a strategic distance from online attacks. 
Currently, learning the Support Vector Machine 
(SVM) computations were used to distinguish port 
sweep efforts based on the new CICIDS 2017 dataset 
with 97.80%, 69.79% accuracy rates were achieved 
separately. SVM may be replaced with alternative 
algorithms like random forest, convolutional neural 
network (CNN), and artificial neural network (ANN), 
which have higher accuracy than SVM (93.29, 63.52, 
99.93, and 99.11, respectively).