Skip to content

Batuu13/Neural-Networks

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Neural-Networks

Parameterized Neural Network Implementation

This project is a simple parameterized neural network implementation. I used this for my Machine Learning Class.

Usage

  • mnist_loader.py is used for loading data.
  • test.py is used for a example for using NN class
  • NN.py is the class for Neural Networks.

In the report, you can find my comment and our results for this implementation.

Imports

  • You need to use numpy, gzip, _pickle and matplotlib.pyplot

Dataset

  • MNIST dataset http://yann.lecun.com/exdb/mnist/ contains handwritten digits and corresponding labels. Dataset contains 60000 training and 10000 test images.
  • Dataset and the code to load it are provided to you. You can donwload them from ftp://cs.hacettepe.edu.tr/pub/dersler/BBM4XX/BBM409_ML/Assignment_3/load/. The dataset is split into three sets: Training (50000 images), test (10000 images)and validation (10000 images). You can use validation set for training (50000+10000 images).
  • When you load the dataset, you’ll see that every image is represented with [784x1] vector and has a label. You’ll use the given representation (do not have to extract new features).

About

Parameterized Neural Network Implementation

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages