Explainable Deep Learning: A Visual Analytics Approach with Transition Matrices
Mathematics, Vol. 12, No. 7, pp. 1024
Abstract
This paper presents a visual analytics approach for explaining deep learning model decisions using transition matrices. The method provides interpretable visualizations that help understand how neural networks process and classify input data.
Citation
Pavlo Radiuk, Oleksander Barmak, Eduard Manziuk, Iurii Krak. "Explainable Deep Learning: A Visual Analytics Approach with Transition Matrices". Mathematics, Vol. 12, No. 7, pp. 1024, 2024.