Skip to content

yepedraza/hm-class-nn

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

22 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Handmade classification neural network

The code is divided in two important parts:

  • In the main.py file a classification dataset was created by gaussian quantiles for two classes.
  • In the neuralNetwork.py the architecture of a neural network with a hidden layer was created using only mathematical resources with the help of the library numpy.

The neural network has some characteristics:

  • Training and validation functions
    • Forward function
    • Back Propagation function (for training)
    • Gradient descent function (for training)
  • Random parameters initialization
  • Hyperbolic tangent activation function
  • Mean Square Error (MSE) loss function
  • All the hyperparameters are worked in dictionaries

Made by @yepedraza

About

Handmade classification neural network

Topics

Resources

License

Stars

Watchers

Forks

Languages