Topological data analysis for detecting anomalies in cyber networks

Cyber networks are complex systems with a highly disorganized global structure consisting of many interconnected units. Graph representations of cyber networks represented by Netflow data and the results obtained with graph algorithms are very important for the security of these networks. This thesis presents a topological data analysis method over simplex complexes established with the help of graph theory for anomaly detection in a cyber network. This method is based on the use of a similarity measure between the barcodes obtained by calculating the persistent homologies of the cyber network. With this measurement, anomalies in the network will be detected and necessary measures can be taken for the current attack situation. In addition, this method, which will be a basis for data analysis, can be used by different disciplines.