The authors extend the work of Niyogi, Smale, and Weinberger on homotopy learning to subsets of Euclidean spaces and Riemannian manifolds with positive reach. They provide tight bounds on the sampling parameters for successful homotopy inference.
The authors extend the work of Niyogi, Smale, and Weinberger on homotopy learning to subsets of Euclidean spaces and Riemannian manifolds with positive reach. They provide tight bounds for the learning process based on sample quality parameters.