This is a simple neural network in Python with single input layer and output neuron.
This is a perceptron type neural network. (A neural network with no hidden layers is called a perceptron.)
In the train_version.py I train the neural network 10000 iterations. Although it's not directly usable , it can be implemented in other own Neural Network as a fundamental program. This will also help you get a basic understanding of how a neural network can be trained and how neural networks work in general.
The program is written entirely in Python and the only library used is Numpy. The neural network uses Sigmoid funtion (Although it prevents the neural network to get 100% accurate answer.)
A simple implementation program (main.py) of the Neural Network is also available in the repository. The main.py tries to predict the output given 3 binary inputs. If the first input is 1, the output should be one. Otherwise the output should be 0.