Sharp Rates in Dependent Learning Theory: Avoiding Sample Size Deflation for the Square Loss

Open in new window