Model Summary
The model summary displays information about the results of training and applying the final network to the holdout sample.
- Sum-of-squares error is displayed because the output layer has scale-dependent variables. This is the error function that the network tries to minimize during training. Note that the sums of squares and all following error values are computed for the rescaled values of the dependent variables.
- The relative error for each scale-dependent variable is the ratio of the sum-of-squares error for the dependent variable to the sum-of-squares error for the "null" model, in which the mean value of the dependent variable is used as the predicted value for each case. There appears to be more error in the predictions of Length of stay than in Treatment costs.
- The average overall error is the ratio of
the sum-of-squares error for all dependent variables to the sum-of-squares
error for the "null" model, in which the mean values of the dependent
variables are used as the predicted values for each case. In this
example, the average overall error happens to be close to the average
of the relative errors, but this will not always be the case.
The average overall relative error and relative errors are fairly constant across the training, testing, and holdout samples, which gives you some confidence that the model is not overtrained and that the error in future cases scored by the network will be close to the error reported in this table.
- The estimation algorithm stopped because the error did not decrease after a step in the algorithm.