Export Reviews, Discussions, Author Feedback and Meta-Reviews

The rebuttal from the authors was concise. However, I was not convinced about the assumption in Eq. 14 of the paper and how the authors defended it. The authors say: "Many documents (text categorization),[..] or time series signals (speech recognition) in a training set are alike. This fact is not systematically exploited by any existing stochastic optimization method!" I don't think this is correct.