This is a simple project where I’ve built an Artificial Neural Network (ANN) from scratch to recognize handwritten digits (0–9) using the MNIST dataset..
- ✅ Custom ANN implementation using NumPy
- 📊 Trained on MNIST (60,000 training + 10,000 test images)
- 🧪 Visualization of predictions and accuracy
DigitRecognitionUsingNeuralNetwork/
├── digit_recognistion_using_Neural_network.ipynbFull implementation in Jupyter
├── check.py Script to test predictions
├── README.md This file
- Python 3.x
- NumPy
- Matplotlib
- Jupyter Notebook
# Clone the repository
git clone https://github.com/AyushPrakash414/DigitRecognitionUsingNeuralNetwork.git
cd DigitRecognitionUsingNeuralNetwork
# Install dependencies
pip install numpy matplotlib
# Run the notebook
jupyter notebook digit_recognistion_using_Neural_network.ipynb
# 📊 Sample Output
✅ After training, the model reaches around 92% accuracy
You’ll see prediction samples like:
Prediction: 3 | Actual: 3
Prediction: 7 | Actual: 7
Prediction: 1 | Actual: 1
You can also save and display some digit images with their predicted labels...