Welcome to neuralpy’s documentation!¶
This page desribes how to get set up with neuralpy
and how to start using it.
We always need help! Contribute on Github!
neuralpy
is a machine learning library specializing in neural networks. It’s intuitive interface
makes it easy for you to quickly start training data and testing models.
The newest package features:
- simple processes for loading, training, and testing data
- variable input and output sizes.
- flexibility with various network structures, activation/output functions, and layer connectivity.
- graphical cost and epoch surveillance through
matplotlib
. - variable mini-batch sizes for stochastic gradient descent.
- abstraction and modularization of backpropagation.
- extensibility for developers to add their own features, layer types and activation functions.