SEARCH RESULT

The technology is growing rapidly and cloud computing usage is increasing. Most of the big and small companies use the cloud nowadays. Cloud computing has the economic benefit which is paid as you use (i.e. pay on the demand). With the increase of cloud usage, security problems on the cloud also increasing. Some mechanisms like firewall, vulnerability scanners and Intrusion Detection System (IDS) and other methods are used to mitigate the intrusions, but they are not enough to detect attacks against the cloud due to new intrusion releases. There are a variety of security methods for improving cloud security from threats and vulnerabilities. In this paper, a new hybrid cloud-based IDRS based on Grey wolf optimizer (GWO) and Neural Network (NN) is proposed to secure and detect intrusions over the cloud. GWO is one of the effective metaheuristic algorithms in many fields such as security. In this paper, GWO is employed to train an NN and the results are compared with other classification algorithms. For experimental results, most up-to-date intrusion

International Conference on Cyber Security and Computer Science
ICONCS

İsmai M. Nur Erkan Ülker

Due to improving technology and spreading internet the entire world, people adapted using it in an extensive manner. Our critical private data are encountering threads which are coming outside of the computer systems and network environments. In other word, intruders access folk’s information without authentication and unauthorized mode. To overcome such kind of security vulnerability matter, a lot of scientific researchers have attracted their awareness the use of this new model called hybrid intrusion detection systems, which is an integration of two or more algorithms, then one of the algorithms is utilized as input while the functionality of other one is tasking or classifying. The new model has a very powerful and plays a significant role in cybersecurity. In recent years, the combination of machine learning methods with metaheuristic algorithms is hybridized to obtain an optimum solution. In this study, we present a new model using the Sine Cosine Algorithm for feature selection and the Naïve Bayes Classifier (NBC) algorithm for classification. Our main goal is to find a model that emphasizes a good performance for detecting and finding preferable accuracy. We compare our experimental results with other algorithms such as KNN, Decision Tree classifications and etc., to realize which one is performed an excellent in terms of accuracy and detection rate. İn addition to this, Sine Cosine Algorithm will be contrast to Particle swarm optimization (PSO), not only PSO but also genetic algorithm (GA) in terms of feature reduction and selection the quality ones, various datasets such as NSL-KDD, ISCX 2012 and etc., has been applied on the new presented method to examine its performance. Finally, the introduced method will prove that whether it has better performance and superior accuracy compared to the other algorithms.

International Conference on Cyber Security and Computer Science
ICONCS

SALAAD MOHAMED SALAAD Erkan Ülker