Skip to content

skogly/numpy-net

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

NumPy Neural Network

This is a Neural Network written in Python by using the NumPy library.

The model implements backpropagation and Stochastic Gradient Descent (SGD) to learn simple datasets.

Important dependencies are NumPy, Scikit-Learn, Pandas and Matplotlib. The exact versions used in this repo are found in requirements.txt, however, this is an auto-generated file from conda on platform osx-arm64 (Mac M1).

The screenshots below are the results from a neural network with 64 nodes in one hidden layer.

Breast Cancer dataset

Accuracies Losses
image image
Predictions Target
image image
5-fold Cross Validation Confusion Matrix
image image

Iris dataset

Accuracies Losses
image image
Predictions Target
image image
5-fold Cross Validation Confusion Matrix
image image

Moons dataset

Accuracies Losses
image image
Predictions Target
image image
5-fold Cross Validation Confusion Matrix
image image

Wine dataset

Accuracies Losses
image image
Predictions Target
image image
5-fold Cross Validation Confusion Matrix
image image

Digits dataset

Accuracies Losses
image image
5-fold Cross Validation Confusion Matrix
image image

About

Neural network written in Python. Implements backpropagation and Stochastic gradient descent (SGD) to solve toy datasets

Topics

Resources

License

Stars

0 stars

Watchers

1 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages