Persistent Homology Based Computational Data Analysis On Diabetes Mellitus

Abstract

The recent evolution in the dimensionality and volume of data makes it more difficult to examine and comprehend the findings. To encounter this, we use Topological Data Analysis (TDA), a paradigm for data analysis that identifies the geometrical structure of data and provides an interpretation based on the data's topological properties. In this paper, persistence diagrams of data from female patients with and without diabetes are computed and compared using bottleneck distance and Wasserstein distance to identify the less significant attributes that lead to the development of diabetes mellitus.