Towards best practice in explaining neural network decisions with LRP

Open in new window