Multilayer-precertron

An implementation of 2 hidden layer neural network using numpy and mnist dataset from tensorflow.

The Multilayer preceptron consist of 1 input layer, 2 hidden layer and 1 output layer.

Input layer = 784 neurons;

1-hidden layer = 256 neurons; relu {activation function};

2-hidden layer = 256 neurons; relu {activation function};

output layer = 10 neurons; softmax {activation function}; cross entropy {loss function};

weight-optimization = adam;

Taken from tensorflow example is at link: https://github.com/aymericdamien/TensorFlow-Examples/blob/master/examples/3_NeuralNetworks/multilayer_perceptron.py