Using cluster analysis techniques to optimize the qualitative assessment of information security risk

The study is devoted to the accuracy of information security risk assessment. The paper substantiates the relevance of risk assessment, based on the consequences of their implementation for business and the probability of their occurrence. The method of qualitative assessment of information security risks (the method of expert assessment) is analysed on a specific example. The application of cluster analysis methods is justified. In detail, the examples show the use of cluster analysis methods: the nearest neighbor method; the remote neighbor method; the k-means method. The principal disadvantages of the first two methods are: the appearance of large clusters that do not have similarities; the lack of the ability of experts to set the desired number of clusters in advance. The application of the k-means method is justified - the ability of experts to set the desired number of clusters in advance by setting the initial centers. The results obtained with the usual qualitative assessment are compared with the results obtained by the methods of cluster analysis. The expediency of using cluster analysis methods to improve the accuracy of information security risk assessment is justified.