Detection of Space-Time Disease Clusters Using A Matrix Factorization Method
Space-time cluster detection has important applications in public health management and epidemiology to devise disease prevention strategies and to find the causes of a particular disease outbreak in a country. This study introduced a new method to detect the potential space-time clusters with no restriction on cluster shape and size and further visualize them distinctly on the heat map. The proposed algorithm is based on matrix factorization technique to find the significant components in spatial as well as temporal dimension. Applications to malaria data in Khyber Pakhtunkhwa, Pakistan shows that the proposed method is effective in detecting the potential clusters.