Enhancing Cyber Security: A Comparative Study of Artificial Neural Networks (ANN) and Machine Learning for Improved Network Vulnerability Detection

Preview this chapter:

Enhancing Cyber Security: A Comparative Study of Artificial Neural Networks (ANN) and Machine Learning for Improved Network Vulnerability Detection, Page 1 of 1

< Previous page | Next page > /docserver/preview/fulltext/9789815256680/chapter-8-1.gif

As we rely more on the internet in our daily lives, network attacks pose a severe threat to the safety of computer systems and networks. Cybercriminals utilize a variety of methods to access sensitive data without authorization by taking advantage of network flaws. Firewalls and intrusion detection systems, which are common security measures, have not been found to be effective in preventing network attacks. The connection between network vulnerability detection and realizing sustainable development goals lies in the broader impact of cybersecurity on the economic, social, and environmental aspects of sustainable development. Deep learning, a kind of machine learning that makes use of neural networks with numerous layers to understand complicated patterns in data, has attracted the attention of researchers and practitioners in an effort to combat the sophistication of cyberattacks that are becoming more sophisticated. Deep learning has demonstrated potential in identifying network attacks due to its ability to automatically extract features from unprocessed data, enhancing its ability to identify previously unknown assaults.These patterns may include unusually high or low levels of network traffic, adjustments to communication patterns, and other aberrant behavior that could be a sign of an impending attack. The first step in using a deep learning algorithm to detect network attacks is to collect and pre-process the data. In this analysis, we used the NSL-KDD dataset, a freely accessible dataset that includes information on both regular and attack traffic. We can start training our deep learning model once we get the data.Large amounts of data are fed into the algorithm during training, and the neural network's parameters are changed to reduce the discrepancy between expected and actual results. Different deep learning architectures, such as convolutional neural networks (CNNs), are available to us. We can use the model to categorize fresh instances of network traffic as either benign or harmful after it has been trained. The model's performance can then be assessed using measures like recall, accuracy, and precision. Our test findings demonstrate that the suggested deep learning technique works better at identifying network assaults than conventional machine learning algorithms. Deep learning algorithms are better equipped to manage the complexity of network traffic data and extract useful information from it.The security of computer systems and networks is seriously threatened by network attacks, and conventional security measures have not been successful in thwarting them. Our paper offers a thorough manual on how to detect network assaults using deep learning, which has shown potential in this area.