christianThardy / Feed-Forward-Artificial-Neural-Network

A four-layer feed forward artificial neural network from scratch in 100 lines of Tensorflow

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Feed Forward Artificial Neural Network

An 88% accuracy was achieved with a 1e-1 learning rate when changing the activation function from sigmoid to tanh and training on a higher number of epochs.

Reference:

Bridle. Probabilistic Interpretation of Feedforward Classification Network Outputs, with Relationships to Statistical Pattern Recognition. 1989.

About

A four-layer feed forward artificial neural network from scratch in 100 lines of Tensorflow


Languages

Language:Python 100.0%