Anomaly detection on a real-time server using decision trees step by step procedure

Anomaly detection is the process of finding outlying records from a given data set. The aim of this paper is to study a well-known anomaly detection technique on the “Short Message Service Centre” server, used in the telecommunications field to handle and store messages. This server was studied in details, a script was written to gather all the required data that went through a cleaning phase and a labeling process after being deeply analyzed. The decision tree algorithm was chosen to be implemented, and it originally gave a precision of 98.82%. After performing different types of tuning and optimization, the precision reached 99.38% with an effective accuracy of 99.98%. Our approach proved that the application on this type of servers is efficient and leads to very good results, which can also be improved in future studies.