Comments on “An Initial Study on the Relationship Between Meta Features
of Dataset and the Initialization of NNRW”
Abstract
This paper comments on our recently published conference paper entitled
“An Initial Study on the Relationship Between Meta Features of Dataset
and the Initialization of NNRW”.
We point out that the above-mentioned article has a typographical error
in proving that using Gamma distribution to initialize NNRW is not a
good choice, and give the corresponding correct proof.