Towards explainable artificial intelligence for the leukemia subtype recognition

Abstract

In this work, we provide a solution to the leukemia subtype recognition problem. Most approaches used for solving a variety of pattern recognition problems have a drawback: in general, they lack explainability. In this paper, we provide a solution for facing this situation. We describe a model whose stages allow deriving knowledge for solving the leukemia subtype recognition problem, are intelligible for the user. Results show that multiclass recognition is achieved from the solutions obtained by the model through multiple runs.