Sayed Hossein Hashemi - Department of Nuclear Engineering, Science and Research Branch, Islamic Azad university, Tehran, Iran
Sayed Mohsen Hashemi - Sama Technical and Vocational Training College, Islamic Azad University, branch soosangerd, soosangerd, Iran
Aref Sayahi - Sama Technical and Vocational Training College, Islamic Azad University, branch soosangerd, soosangerd, Iran

Today, due to high data volumes and their complexity, we need the right tool for the analysis of existing data and acquiring knowledge. The interest of attains hidden knowledge in Data mining is Data Growth. Data mining in recent years, has significant impact on the academic and industrial environments and found many applications in different fields. This paper presents an integrated approach to data mining to intrusion detection in computer networks by supervised machine learning algorithms that decision trees are the most important approaches. The results showed that there is a comparison between the performance of this algorithm and other methods Therefore, to increase efficiency and reduce the error rate of the algorithm we used adaptive boost classifier combination and reduce the size of the 15 features, the accuracy reached 96.74%, a higher percentage than previous methods.

attack, Intrusion Detection, Data mining, Classification, machine learning