×
The submission system is temporarily under maintenance. Please send your manuscripts to
Go to Editorial ManagerOur proposed method used to overcome the drawbacks of computing values parameters in the mountain algorithm to image clustering. All existing clustering algorithms are required values of parameters to starting the clustering process such as these algorithms have a big problem in computing parameters. One of the famous clustering is a mountain algorithm that gives expected number of clusters, we presented in this paper a new modification of mountain clustering called Spatial Modification in the Parameters of Mountain Image Clustering Algorithm. This modification in the spatial information of image by taking a window mask for each center pixel value to compute distance between pixel and neighborhood for estimation the values of parameters ?, ? that gives a potential optimum number of clusters requiring in image segmentation process. Our experiments show ability the proposed algorithm in image brain segmentation with a quality in the large data sets
The goal of the study is to predict the Man in the Middle attack in the packets of Wireshark program by using Support Vector Machines (SVM).In the time of using the internet, it has become a tool targeted by attackers and hackers; it is a serious threat to the devices. A uniqueness of an attack that appears in multiple identities for legitimate agencies. It is very necessary to know the behavior attack and predict the possible actions of an attacker. In this research a detection of Man in the Middle attack by monitoring the Wireshark program and recording any changes can be recognized in packet information. The classification of packets is divided into two categories (normal and abnormal). The proposed model is designed in many stages: loading data, processing data, training data, and testing data. The detection of SVM based on abnormal network packet through movement packets in the Wireshark program that needs to deal with current packets to recognize a new attack that one does not have prior knowledge of its detection, and there is a need for an intelligent way to separate network packets that represent normal. The proposed approach achieved an accuracy of 97.34% in detecting attacks. The results show that the proposed model effectively visualizes attacker behavior from data that represents abnormal network attackers. Research achieves successful accuracy in predicting abnormalities.