Define Architecture for Handwritten Digit Recognition Neural Network

[Gdsc-Cummins]

Issue: Define Architecture for Handwritten Digit Recognition Neural Network

Description:

We are looking for contributors to define the architecture of a neural network that can recognize handwritten digits as part of our open-source project. The code for the MNIST dataset and neural network implementation can be found in the "MNIST" folder within the "level-medium" directory of our repository.

Task:

Your task is to define the architecture of the neural network, including the number of layers, the type of layers (e.g., convolutional, fully connected), and the hyperparameters (e.g., learning rate, batch size) required for training.

Steps to Contribute:

1. Visit the "level-medium" directory in our repository and navigate to the "MNIST" folder to access the existing codebase and dataset.

2. Review the code to understand the existing structure and dataset.

3. Define the architecture of the neural network in a clear and well-documented manner. Consider factors such as model complexity, layer sizes, activation functions, and any necessary pre-processing steps.

4. Submit your solution by creating a Pull Request (PR) to the repository. Please provide detailed explanations and comments in your code to help reviewers understand your approach.

Review and Merging:

The review of this issue will be done at the end of October. The best solution, based on accuracy and efficiency, will be selected for merging into the project. Make sure to submit your PR before the deadline for consideration.

We appreciate your contributions to this project. Handwritten digit recognition is a fundamental problem in machine learning, and your efforts will help improve our neural network's performance. Thank you for participating, and we look forward to your contributions! Happy coding!

[OkenHaha]

Hey can I contribute to this also? This seems intresting to tackle on