Fig. 5From: Medication adherence prediction through temporal modelling in cardiovascular disease managementAn unrolled view of RNN across the time-step dimension. Here, RNN can be a layer of Simple RNN or LSTM. NN is a layer of densely connected NN with softmax activation. \(x_{n}\) are the inputs across n time-steps. \(\hat{y}\) is the outputBack to article page