Skip to content

Implementing a Neural Network from scratch, exercise

Notifications You must be signed in to change notification settings

r3lativo/nn_experiment

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 

Repository files navigation

Implementing a Neural Network from Scratch - Experimenting

This experiment is based on this blog post and is a part of a report for a ML for NLP course at the University of Trento, CIMeC.

Experimenting with it, I managed to:

  • implement a slightly bigger moon-shaped dataset;
  • divide it into train and test splits;
  • change the gradient descent for a minibatch gradient descent with batch_size = 32;
  • implement an annealing schedule for the gradient descent learning rate, decreasing it each 2000 passages;
  • experiment with other activation function;
  • test each model's accuracy;
  • plot all the results.

Moreover, most of the function have been slighly modified to integrate more modulability (e.g., using the specified activation function, loading the the training dataset...).

About

Implementing a Neural Network from scratch, exercise

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published