Project Submission

Submit Your MNIST Digit Recognition Project

Create a public GitHub repository with the exact name shown below, add all required files, and enter your GitHub username to submit.

Repository Requirements

Your repository must have this exact name:

github.com/<your-username>/mnist-digit-recognition
Required Files and Folders
  • notebooks/
  • mnist_cnn.ipynb
  • models/
  • best_model.h5
  • reports/
  • confusion_matrix.png
  • training_history.png
  • requirements.txt
  • README.md
README.md Must Include:
  • Your full name and submission date
  • Final test accuracy achieved (target: over 99%)
  • CNN architecture summary and layer descriptions
  • Image augmentation techniques used
  • Training curves (accuracy and loss plots)
  • Confusion matrix visualization
  • Key findings and error analysis
Required (submission will fail without these)
Dataset Reference

The MNIST dataset should be loaded from TensorFlow/Keras or downloaded from Kaggle. Do NOT include the dataset files in your repository (they are too large).

MNIST on Kaggle MNIST CSV Format
Dataset Info: 70,000 images | 28x28 pixels | Grayscale | 10 classes (0-9) | Load via tensorflow.keras.datasets.mnist

Submission Form

github.com/
Enter only your username. We will verify the repository mnist-digit-recognition exists in your account.

Your repository will be verified before submission is accepted

Project Submitted Successfully!

Your project has been verified and submitted. Great work on your MNIST Digit Recognition project!

Back to Project Continue Course