- Neural networks and backpropagation (2012) – Good post with full mathematical derivation and accompanied GitHub repository.
- Automatic differentiantion
- Python code for NN and notebook
- Tensor talk – latest AI code
- How I learned to code Neural Network – a good post with lots of references to resources for learning NN and another reference to the author’s data-sets site. Some pointers:
- View Welch labs on Youtube and week 4 of Andrew Ng course.
- Decide a simple data set for experimentation
- Refresh your logistic regression skills (week 3 of Andrew Ng course)
- Understand back-propagation:
Examples for coding complete NN in Python
- I am trask blog – incudes NN basics, backpropagation, gradient descent, dropout, LSTM-RNN. Short examples with explanatory purposes. Do not aim at writing complete NN implementation
- Training (deep) neural network – blog post by Upul Bandarea. Usethe MNIST data set for recognizing handwritten digits.
- Neural networks and deep learning – hand-on book by Michael Nielesen which works on recognizing handwritten digits.
- CS231n – CNN (Stanford course) – Good course notes in here. It is possible to use the ipython notebooks to do the programming assignments. The code use CIFAR classification problem and gives a lot of pre-written code, e.g. to read and display the data sets. The coding include writing specific parts of the code focusing on efficient vectorized implementation.