https://github.com/keyurparalkar/mnist-digit-recognizer

Learning computer vision fundamentals with the MNIST data
https://github.com/keyurparalkar/mnist-digit-recognizer

cnn deep-learning kaggle-digit-recognizer lenet-architecture mnist

Last synced: 2 months ago
JSON representation

Learning computer vision fundamentals with the MNIST data

• Host: GitHub
• URL: https://github.com/keyurparalkar/mnist-digit-recognizer
• Owner: keyurparalkar
• Created: 2018-07-27T14:35:46.000Z (almost 7 years ago)
• Default Branch: master
• Last Pushed: 2018-07-27T17:15:11.000Z (almost 7 years ago)
• Last Synced: 2024-12-30T19:29:19.599Z (4 months ago)
• Topics: cnn, deep-learning, kaggle-digit-recognizer, lenet-architecture, mnist
• Language: Jupyter Notebook
• Homepage: https://www.kaggle.com/c/digit-recognizer
• Size: 151 KB
• Stars: 0
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
## MNIST-Digit-Recognizer

### Dataset information

MNIST ("Modified National Institute of Standards and Technology") is the de facto “hello world” dataset of computer vision. Since its release in 1999, this classic dataset of handwritten images has served as the basis for benchmarking classification algorithms. As new machine learning techniques emerge, MNIST remains a reliable resource for researchers and learners alike.

### Aim

Goal is to correctly identify digits from a dataset of tens of thousands of handwritten images.

### Model

Creating LeNet5 nn class module

    conv2d => relu => maxpooling => conv2d => relu => maxpooling => fully connected layer(fc)1 => fc2 => softmax output

### Libraries required

    pytorch

    numpy

    pandas

    matplotlib

    

### For installing pytorch

    conda install pytorch torchvision -c pytorch

### For running this project

1. Activate the fastai environment: `source activate pytorch`

2. Run the notebook: `jupyter notebook`