Greeting. I am Chin Luh, the author for Neural Network module in atoms
portal (https://atoms.scilab.org/toolboxes/neuralnetwork/2.0). Thlde
module is base on the book "Neural Network Design" by Hagan Martin (1st
version) and the codes are following the pseudo code given in the book.
If you are interested, you could look into this module and implement the
"Regularization" into the algorithms for all the FFBP in the module,
which is highlighted in one of the session in Andrew Ng's machine
One thing you might need to be careful: In Machine Learning Course you
attended, Prof Andrew using the notation where the training dataset
arranged in rows, and the features arranged in columns. This is the
reverse with this module, the book this module referred to, as well as
the standard way python used. However, In Prof Andrew first course in
Deep Learning specialization, he used the same notation with this module.
As this module is totally in Scilab code, I personally believe that it
is easy to start.