An simple project by using the Fashion MNIST data set and neural networks to predict the category of the clothes.
A simple neural network with three leayers, each one of input size 28*28, 100 and 10. The first two using ReLU activation and the last one using linear activation, with from_logits=True
while compiling.
{'accuracy': 0.8525999784469604, 'loss': 0.4540776014328003}
Need further improvements afterwards.