The operations of cyber-attacks are expanding due to the increasing use of the Internet networks, which are considered a "significant" threat to the security of the information sent and received in the event of a cyber-attack that mainly threatens many confidential information as well and causes a DoS denial of service. The goals of cyber-attack detection are to identify the type of cyber-attack to provide additional intensive insight into application designers as well as ongoing demands for security. Where content creators operate a developed identification framework that can distinguish from the start attacks like DoS, malware etc. In this thesis, an approach to detect DoS attacks is presented. The proposed approach tries to provide a targeted framework for providing an Intrusion Detection System (IDS) using a combination of machine learning and evolutionary algorithms. Since in these types of systems, datasets have a serious role in the system creation process, as a result, we will refine the dataset in the first step. In this step, using the genetic algorithm, we selected a desirable subset of features so that we can reduce the complexity of the data set. The second step of the proposed method includes a data segmentation algorithm in which the sample data is divided into segments with similar characteristics. For this process, we have taken the help of Artificial Bee Colony (ABC) which is enhanced with Haar Transform objective function. This step will help to classify the data. In the last step of the proposed method, a decision tree is used to classify healthy DoS sessions. In order to evaluate the proposed method, a benchmark dataset called NSL-KDD has been used. In addition to the proposed method, one of the latest articles in the research field of this thesis has also been implemented and evaluated using the MATLAB language. The results of evaluations in the form of Precision, Recall, Accuracy and F-Measure criteria and different scenarios show that the proposed method has been able to provide favorable results
