A Machine Learning Based Approach for Anomaly Detection

Abstract

The amount of data accumulated in the databases of the companies increases over time. For this reason, the difficulty of using these data by the people and systems is increasing. Furthermore, it would not be wrong to say that it has become more difficult to ensure the security of data. Intrusion attempts or attacks on such systems must be prevented. The process of determining the events or observations that occur on the data and that arouse suspicion is known as intrusion detection systems (IDSs). Anomaly detection studies are also among IDS methods and used for the same purpose In this study we tried to build a machine learning model which is based on artificial neural network algorithm. CRISP-DM analytics model was followed which is a widely used standard process for data mining. Encouraging results were obtained with the proposed method. The results were compared with other methods in the literature.