This is a simple implementation of a neural network using Numpy. In prepare_data.py a training and test data set is prepared. The current algorithm is trained to classify 3x1 tensors where the first entry is 10. The user should feel free to use his or her own input data, and edit the code accordingly. In neural_network.py the neural network class and its functions is defined. The class includes functions for forward and backward pass, neural network training, and model evaluation.