I am trying to understand the contribution of different time series to a multi-step ahead neural network forecasting model. The shape of the input vector is (n_samples, n_timestamps_input, n_features) and the shape of the prediction is (n_samples, n_timestamps_output). I would like to understand how the different i input samples -> (n_samples, n_timestamps_input, i) contribute to the prediction j -> (j, n_timestamps_output). Any advice?
Asked
Active
Viewed 98 times
