updated CNN and DNN unit tests

[kamuda1]

This PR updates the CNN and DNN unit tests for the Tensorflow classes in the model_classes.py module. Tests use fixtures to create models with dense connections and convolutional filters of all ones. Training tests and tests for the CAE and DAE to come.

Tests include checking if:

• dropout changes network output when applied (during training)
• dropout changes network output when not applied (during inference)
• l2 regularization adds to the cost function
• output values to unit inputs (inputs of all ones) make sense

[pep8speaks]

Hello @kamuda1! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

There are currently no PEP 8 issues detected in this Pull Request. Cheers! :beers:

Comment last updated at 2019-09-18 23:33:30 UTC

[samgdotson]

Everything looks good, merging PR