Routing and Scheduling for Mobile Vaccination Services Aiming Fairness for the Vulnerable Populations

There are 225 neighborhoods in the Şehitkamil and Şahinbey districts of the city Gaziantep in Turkey. The distance matrix of these 225 neighborhoods is given in meters in the data set. We created small problem instances by grouping our locations together. The clustering task being performed utilizes the k-means algorithm from the sci-kit-learn library in Python to cluster 225 locations based on their longitudes and latitudes. The algorithm groups similar locations together by iteratively finding the mean of the points in each cluster, and then reassigning points to the cluster whose mean is closest to them. In this analysis, clustering is performed with 8, 10, 12, and 20 clusters to evaluate the optimal number of clusters for the data. Once the clustering is complete, the results of the analysis will include the assignment of each location to a specific cluster as well as the coordinates of the centroid for each cluster. This information can be used to gain insights for assigning the demand of all points in the same cluster and having one point representing the cluster. we used the centroid coordinates to provide a convenient representor of the demand location of each cluster.