Reviews: Integrated accounts of behavioral and neuroimaging data using flexible recurrent neural network models

In this paper the authors use a recurrent neural network model to study reward processing in human fMRI experiments. The main novelty of this approach is that it takes into account both the neural data (measured by fMRI) and the behavioral outcomes. In contrast the standard approach is to fit a computational model on the behavioral data, then use that as a regressor in a classical GLM analysis, which may find a good solution for the behavioral data but one that is different from the solution implemented in the brain. In this respect, this application stands out. Most other neural network models that are used to model fMRI data (e.g.