K-means clustering as a tool for effective distribution of resources and policy making: Using Nigeria Covid-19 medical supplies as a case study

In this study, K-means is one of the non-hierarchical clustering methods which was used to assess how the Covid-19 medical supplies donated to Nigeria were distributed across the States. Secondary data on the medical supplies donated and distributed were collected from the Nigerian Centre for Disease Control (NCDC) website. On the average, 2692.92, 270.243 and 29.7297 face masks, coverall gowns and face shields were distributed across the 36 States and Federal Capital Territory of Nigeria. On applying the K-means clustering method using the K-means++ technique in Python software (Anaconda 3-5.2.0), we obtain 4 given as 0, 1, 2 and 3 with each cluster having 9, 1, 26 and 1 state, respectively, for Covid-19 cases. As for the medical supplies, an inertia value of 8.84 was obtained and number of cluster was still taken as above. The cluster 0, 1, 2 and 3 had 34, 1, 1, and 1 state, respectively. From the result of clustering, there are indications of uneven distribution of the medical supplies based on Covid-19 cases following the number of states grouped into different clusters. Based on these observations, to ensure even distribution of resources based on the prevalence of Covid-19, Plateau, Oyo, Rivers, Edo, Kaduna, Delta, Kano and Ondo should receive medical supplies equivalent to that of Ogun. We present an optimum and efficient means of distributing resources especial related Covid-19 supplies in the future.