1 paper
Mohammadhossein Toutiaee, John Miller
We propose a new framework for 2-D interpreting (features and samples) black-box machine learning models via a metamodeling technique, by which we study the output and input relati…