tsarjak / mnist-basic-neural-net

Basic neural network (Fully connected) with three hidden layers using tensorflow for python for classifying digits from MNIST data sets with 95% accuracy.

Basic Neural Network (Fully Connected) for MNIST

Basic neural network (Fully connected) with three hidden layers using tensorflow for python for classifying digits from MNIST data sets with ~95% accuracy.

With three hidden layers, containing 100, 50 and 20 neurons respectively, and training over 100 epochs, this network gives approximately 95% accuracy on the test data of the mnist data set.