R help - Feb 2018 - Convolutional neural networks (CNN) - build a model and After?

Hi,
I am learning CNN using MXNet R package. I am following this great tutorial
about olivetti_faces reconnaissance
<https://www.r-bloggers.com/image-recognition-tutorial-in-r-using-deep-convolutional-neural-networks-mxnet-package/>.

In the end after building model and testing the final score was 0.975.
It is great score but what can do after with this model?
How can use this model for other set on images?
Is it used to recognize which person in the training data or classify new
images if it content faces or not?
Thanks
Karim

	[[alternative HTML version deleted]]