https://github.com/caronl/mnist_app

Repository for this app :
https://github.com/caronl/mnist_app

digit-recognition keras shinyapps tensorflow

Last synced: 8 days ago
JSON representation

Repository for this app :

• Host: GitHub
• URL: https://github.com/caronl/mnist_app
• Owner: caronl
• Created: 2017-08-25T02:00:44.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2018-01-07T21:33:44.000Z (almost 7 years ago)
• Last Synced: 2024-08-13T07:14:53.188Z (4 months ago)
• Topics: digit-recognition, keras, shinyapps, tensorflow
• Language: R
• Homepage: https://exhibit-a.shinyapps.io/mnist_app/
• Size: 9.67 MB
• Stars: 3
• Watchers: 1
• Forks: 0
• Open Issues: 3
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• jimsghstars - caronl/mnist_app - Repository for this app : (R)

README

        
# mnist_app

### app.R

This script contains the ui and server sides of the app.

### model.hdf5

This is the model created by train.R (see [src/](#src)).

### src/

This folder contains all the relevant R programs:

- train.R was used to train the model

- assess_holdout.R allows to assess 'real-life' performance of the model (see [holdout](#holdout))

- other files are helpers called by the other scripts

### holdout/

When ran locally, this app allows the user to save the images created by the user, as well as

their corresponding labels as identified by the user. This 'holdout' data can then be used for

validating that the model performs well in a different context than the original MNIST. Indeed,

the size of the digit, the width of the pen and the difference in the way the drawing was

obtained (real pen vs. drawing with a computer mouse) can all affect the 'real-life' performance

of the model.

### www/

This folder contains the GitHub "Fork Me" logo and the javascript signature pad used in the app. Please see

[szimek's GitHub repo](https://github.com/szimek/signature_pad) for the original source code.

### CSS_file.css

This file is necessary for the signature pad to work properly.