I am learning basics of deep learning using python and trying out various algoroithms. I usually use Jupyter notebooks. If you using the material in this repository, i would recommend using Jupyter notebook files. They are better documented and easier to follow. If you are looking for code to execute, I have rough patched the same code in .Py files in the same folder next to the notebooks. Feel free to contribute if you like :)
This folder is all about basics of python. how to time processes, basic methods like sigmoid, Relu etc., scripts to convert images to vectors and very minimal pre-processing.
Did you know, single neuron almost identically map logistic regression. Hence the scripts inside this folder mostly use logistic regression terms.
Hidden Layer
Neurons do not work alone. There are many of them, forming layers . Each layer passes on what it has learnt to the next layer. And each of the next layer gives a feedback to the previous one when it hears back from target.
This is the general L layer NN. This is what all packages like TensorFlow, CNTK etc would provide. Mostly with major improvements. :D