Reviews: Distributed estimation of the inverse Hessian by determinantal averaging

Neural Information Processing Systems 

Additional Comments: • Overall, the article is well-written and structured. It has a clear contribution and also significant theoretical justification. There are only a few mistyping and grammatical errors: Line 17: "tranformation" "transformation" Line 133: " its entries is of …" "its entries are of …" Line 146: " accross" "across" Line 150: " emprical" "empirical" • In line 54, it is better to define in the theorem 2 along with . If there is no constraint on the number of estimators, this means choosing subsamples is with replacement. However, in line 158, the authors claim that their method is without replacement.