There are abundantly many explanation methods for explaining deep neural networks (DNNs), each with its advantages and disadvantages. In most cases, we are interested in local explanation methods, i.e. explanations of the network’s output for a particular input, because DNNs tend to be too complex to be explained globally (independent of an input).

… In this short article, I will present one fundamental attribution technique: occlusion analysis. The basic concept is as simple as they come: For every input dimension of an input x, we evaluate the model with that dimension missing, and observe how the output changes. Read More