Topological analysis of data on 911 calls in the Boston area

Among the newer approaches to data analysis are topological methods (TDA), which proved to be effective in analyzing data. In this thesis we analyze data on 911 calls that include a large number of calls. Firstly, we prepare data by grouping calls together using the Vietoris-Rips complex. We do this because it enables us to also analyze smaller areas and connect them. We analyze this complex in two ways: by using Morse theory and persistent homology. Morse theory is used to acquire critical simplices from the complex. They give us new information about the data. Using persistent homology, we produce persistent diagrams that illustrate how homology of a complex changes depending on a parameter. The initiative to use the TDA on such data came from the Department of Sociology at Harvard, where they had already tried to analyze this data by using various mechanical and mathematical models.