A1688
Title: A robust persistent homology trimming approach
Authors: Subhra Sankar Dhar - IIT Kanpur (India) [presenting]
Tuhin Subhra Mahato - IIT Kanpur (India)
Abstract: A robust version of persistent homology based on trimming methodology is studied to capture geometric features through support of the data in the presence of outliers. The proposed methodology operates when outliers lie outside the main data cloud as well as inside the data cloud. Theoretical analysis establishes that the Bottleneck distance between the proposed robust version of persistent homology and its population analogue can be made arbitrarily small at a certain rate for sufficiently large sample size. The practicality of the methodology is demonstrated on various simulated data and benchmark real data associated with cellular biology.
