Кластеризація регіонів України на основі епідеміологічної ситуації ВІЛ/СНІД

It is a well-known fact that Ukraine is among the countries most affected by the HIV/AIDS epidemic. In this work, I fit a mixed linear model to cluster the regions of Ukraine based on the number of HIV/AIDS diagnoses and AIDS-related deaths. Using this model, I represent each region as a combination of stochastic vectors, each describing the probabilities of a region belonging to a particular class based on the normalized numbers of HIV/AIDS diagnoses and AIDS-related deaths in that region. I then construct a graph where each region is represented as a vertex, and each edge is weighted by the correlation between the corresponding class-belonging vectors. Finally, I perform Walktrap clustering to group regions according to the similarity of HIV/AIDS epidemic trends and interpret the resulting clusters. The aforementioned work is conducted as part of an actuarial study of the HIV/AIDS epidemic in Ukraine.