An example of how to create a simple self-developed Neural Network from scratch in Python
I'm not sure if I have implemented everything correctly. Please tell me (or better: make a PR) if you have found a mistake.
Shows how to use NeuralNetwork.py
Contains the Neural Network using L-BFGS for minimizing the costs.
You can pull the repository the way it is. Running Main.py calls the Neural Network with the breast_cancer dataset. If you just want to use the Neural Network you have to initialize it with the number of hidden nodes (and optionally: number of hidden layers and epsilon for random initialization) and then call fit() with a Numpy Array X (m_samples, n_features) and a Numpy Array y (m_samples).
If you have an idea how to improve this Neural Network keeping it as simple as possible please fork it and make a PR. Like said before: I'm not 100% sure if it is working correctly.