| summary_nn {lazy.nntools} | R Documentation |
Calculation of the Summary
summary_nn(W, X, Y, linear = 1, softmax = 0, allO = 0)
W |
List of Weight Matrices |
X |
The input regressor matrix: the first layer |
Y |
The criterion variable matrix |
linear |
= 0 to apply sigmoid functions to the last layer. |
softmax |
Not Yet Available |
all0 |
= 1 to return the O list |
A list of
n # of observations
np # of the X variables
nq # of the Y variables
nlayer # of layers
nn # of neurons per layer
nparam # of parameters estimsted
rss residual sum of squares
rmse rmse
rmse1 rmse for each Y variable
maxag max absolute gradient
Yhat Yhat matrix: output of the last layer
grad gradient: d rss / unlist(W)