In general, updating a model involves putting a new training set into the model for training. But my question is, in an LSTM with multiple outputs (say, five outputs), I only have new data for one of the outputs, how do I update the model?
I built neural networks model in keras