https://github.com/agamiko/tiny-mnist-challenge
Tiny MNIST Challenge for university student's class with example codes in matlab and pytorch.
https://github.com/agamiko/tiny-mnist-challenge
challenge convolutional-neural-network example-code pytorch student-project toolbox tutorial
Last synced: 7 months ago
JSON representation
Tiny MNIST Challenge for university student's class with example codes in matlab and pytorch.
- Host: GitHub
- URL: https://github.com/agamiko/tiny-mnist-challenge
- Owner: AgaMiko
- License: mit
- Created: 2020-04-04T19:49:59.000Z (almost 6 years ago)
- Default Branch: master
- Last Pushed: 2020-04-23T10:49:09.000Z (almost 6 years ago)
- Last Synced: 2025-06-07T10:48:40.579Z (8 months ago)
- Topics: challenge, convolutional-neural-network, example-code, pytorch, student-project, toolbox, tutorial
- Language: Jupyter Notebook
- Homepage: https://www.kaggle.com/c/msi-tinymnist2020
- Size: 2.95 MB
- Stars: 0
- Watchers: 1
- Forks: 5
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Tiny MNIST Challenge 2020
Tiny MNIST Challenge for Gdańsk University student's class.
Kaggle: https://www.kaggle.com/c/msi-tinymnist2020
### Matlab examples - code and toolboxes instruction.
Built-in data loaders and submission preparation.
* **Simple neural network** - A framework for using matlab toolbox to design basic fully-connected network. [code](example_code/matlab_examples/simple_Network.m)
* **Basic convolutional neural network** - Simple neural network with 2 convolutional layers, relu activation, maxpooling and fully-connected layer.[code](example_code/matlab_examples/simple_CNN.m)
### Pytorch examples - ready to run in Colab
Built-in data loaders, submission preparation and sending.
* **Simple neural network** - This is a basic fully-connected network. Two fully-connected layers and relu activation.[code](example_code/pytoch_examples/simple_Network.ipynb)
* **Basic convolutional neural network** - Simple neural network with 2 convolutional layers, relu activation, maxpooling and fully-connected layer. [code](example_code/pytorch_examples/simple_Network.ipynb)
# Winners solutions
The winning codes can be found in the [notebooks section](notebooks). Number of the file indicates the postion of the solution in the rank.