Controlling False Discovery in Arbitrarily Structured Hypothesis Spaces via Reproducing Kernels

Open in new window