Hypersphere Fitting From Noisy Data Using an EM Algorithm

This letter studies a new expectation maximization (EM) algorithm to solve the problem of circle, sphere and more generally hypersphere fitting. relies on introduction random latent vectors having priori independent von Mises-Fisher distributions defined hypersphere. statistical model leads complete data likelihood whose expected value, conditioned observed data, has Von distribution. As result, inference can be solved with simple EM algorithm. The performance resulting fitting is evaluated for circle