Statistical Modeling of Ransomware Attacks Trends

For the purpose of this study, a dataset is collected from ransomware attacks database, for a six-year period, starting from 2018 to present. The numbers of ransomware attacks for every month and year are counted and visualized in the form of sums. Different linear and nonlinear statistical models are tested and the most reliable trend lines are selected according to criteria of the highest R-squared value (R2). Trend lines for each of the observed years follow the sextic polynomial curve, with the values of R2 from 0.6910 to 0.9646. The analysis of the number of attacked sub-industries is conducted and percentage share by sub-industries for each year is calculated. The results are presented separately in the form of graphs for each year, as well as for the six-year period, indicating that the most affected were healthcare, government, and education. The attitude regarding the appearance and trends in frequency of different ransomware strains through the years is also observed. There is no regularity, but it is possible to draw some other trends, with respect to the number of recorded strains. The results of this study represent a base for future predictions for the number of attacks and ransomware strains.