Coefficient-based regularized distribution regression

In this paper, we consider the coefficient-based regularized distribution regression which aims to regress from probability measures to real-valued responses over a reproducing kernel Hilbert space (RKHS), where the regularization is put on the coefficients and kernels are assumed to be indefinite. The algorithm involves two stages of sampling, the first stage sample consists of distributions and the second stage sample is obtained from these distributions. The asymptotic behavior of the algorithm is comprehensively studied across different regularity ranges of the regression function. Explicit learning rates are derived by using kernel mean embedding and integral operator techniques. We obtain the optimal rates under some mild conditions, which match the one-stage sampled minimax optimal rate. Compared with the kernel methods for distribution regression in existing literature, the algorithm under consideration does not require the kernel to be symmetric or positive semi-definite and hence provides a simple paradigm for designing indefinite kernel methods, which enriches the theme of the distribution regression. To the best of our knowledge, this is the first result for distribution regression with indefinite kernels, and our algorithm can improve the learning performance against saturation effect.