Welcome to the Handwritten Digit Classifier section! This folder contains implementations of neural network models trained to classify handwritten digits from the MNIST dataset. The models covered include Artificial Neural Networks (ANNs) and Convolutional Neural Networks (CNNs), which are widely used in image classification tasks.

Note: The notebooks in this folder provide foundational concepts and practical implementations of ANN and CNN models. For a deeper understanding, additional recommended resources are provided below.

This folder currently includes:

• Artificial Neural Network (ANN): A fully connected neural network for digit classification.
• Convolutional Neural Network (CNN): A deep learning model designed to handle spatial relationships in image data.

Each section includes assignments to reinforce learning, along with solutions for self-assessment.

Follow these steps to build a strong foundation in handwritten digit classification:

• Purpose: Implement a simple fully connected feedforward neural network for digit classification.
• Topics to Cover:
  • Introduction to ANNs
  • Network architecture (input, hidden, and output layers)
  • Activation functions and optimization techniques
  • Training and evaluation of ANN on MNIST
• Resources:
  • Keras ANN Documentation
  • ANN for Digit Recognition
  • Deep Learning with ANN (Video)

• Purpose: Train a CNN model designed to extract spatial features from images and improve classification accuracy.
• Topics to Cover:
  • Basics of CNNs (convolution, pooling, and fully connected layers)
  • Building a CNN architecture for MNIST
  • Improving performance with dropout and batch normalization
  • Training and evaluation of CNN on MNIST
• Resources:
  • Keras CNN Documentation
  • Deep Learning for Computer Vision (Stanford Course)
  • CNN Explanation (Video)

Each section includes hands-on assignments to help solidify your understanding of the concepts. Solutions are provided for self-assessment. It is recommended to attempt the assignments before referring to the solutions.

1. Start with ANN: Learn how a basic fully connected neural network classifies handwritten digits.
2. Move to CNN: Explore how convolutional layers improve performance by capturing spatial patterns.

Happy coding! Mastering these models will enhance your ability to work with deep learning for image classification. For further learning, refer to the documentation and tutorials linked above.