Nonparametric regression in nonstandard spaces

A nonparametric regression setting is considered with a real-valued covariate and responses from metric space. One may approach this via Fréchet regression, where the value of function at each point estimated mean calculated an objective function. second geodesic which builds upon fitting geodesics to observations by least squares method. These approaches are applied transform two most important estimators in statistics – local linear estimator orthogonal series projection estimator. The resulting procedures consist known as well new methods. We investigate their rates convergence general compare performance simulation study on sphere.