Additional input. The output from the dropout is concatenated with the extra input from the demographical information corresponding to the participant going through the sequence of events. This concatenated output is passed on to three fully connected dense layers. The number of neurons in these layers, as well as prior layers, is a hyperparameter which needs to be tuned. The last layer is a single output neuron with a sigmoid activation function that outputs a probability for the participant to login to their personal online page after a set of events.
Appears in 3 contracts
Sources: End User Agreement, End User Agreement, End User Agreement