ML/DL/RL templates with very basic datasets with a bit of modification in order to further demonstrate someone else. Can be a good starter CODE!
FOR ANN SECTION: important research papers and links :
https://stats.stackexchange.com/questions/154879/a-list-of-cost-functions-used-in-neural-networks-alongside-applications - differnt type of cost functions!
http://yann.lecun.com/exdb/publis/pdf/lecun-98b.pdf - backpropagation beginning.
https://iamtrask.github.io/2015/07/27/python-network-part2/ - gradient descent!
Installing Tensorflow/theano/keras provided you are working with anaconda and python 3.5.2! so here it goes
-Below are the installation instructions (you will notice that only Keras needs to be installed, this is normal, since the Keras package integrates TensorFlow):
if() Linux and Mac users, please open your terminal. On Mac, the easiest way to open it is to press anywhere cmd + space, and then in the Spotlight Search you enter "terminal". On Linux, you will find it very easily, usually on the left side of your monitor. Then inside the terminal, copy-paste and enter the following line command:
-conda install -c conda-forge keras
else if() Windows users, please open the anaconda prompt, which you can find this way: Windows Button in the lower left corner -> List of programs -> anaconda -> anaconda prompt Then inside the anaconda prompt, copy-paste and enter the following line command:
-conda install -c conda-forge keras
install keras packages with pip, like:
-pip install keras-applications
-pip install keras-preprocessing
And you are ready to use and train YOUR NEURAL NETS!
IMPORTANT read next:
1 https://stats.stackexchange.com/questions/226923/why-do-we-use-relu-in-neural-networks-and-how-do-we-use-it
1 Answers the question, "why do we use linear activation functions like ReLU when working with CNN?"