A Nyström Estimator Error bound

Assumption 2. There exists a positive constant K such that sup Now let us derivative the inference under the introduce of inexact kernel estimations. It remains to estimate the error for estimating the information gain. Now we can prove our main theorem 1. MMD estimator are empirically selected via experiments. SKL-UCB and ERBF-UCB fail to control the ball's ending position under input randomness. Moreover, all the ball's ending positions are well controlled and