A Python package developed to understand the basics of the back-propagation algorithm on toy problems.
The package provides an implementation of the back-propagation approach proposed in: Rumelhart, D. E., Hinton, G. E., & Williams, R. J. (1986). Learning representations by back-propagating errors. Nature, 323(6088), 533-536.
The toy problems include an exclusive-or (XOR) gate and the Rosenbrock function.
TODO: Add dependencies. E.g.:
- Python 2 or 3
- matplotlib
- etc.
TODO: Add usage instructions. E.g.: Run the following code:
python RosenbrokFunction.py
TODO: Show example results