Read more in posts: 1, 2

Video demonstration of mlp training process on the simple example: https://www.youtube.com/watch?v=sYe_gty4R5w

• nn.py
  Contains the implementation of a neural network built from scratch.

• nn_extractor.py
  A decorator class that supplements the NeuralNetwork class with additional methods, such as extracting weights and other parameters.

• draw_digit.py
  A script providing an interactive interface for predicting hand-drawn digit images in real-time.

• mlp_training.ipynb
  A Jupyter notebook that explains the feedforward and backpropagation processes. It imports nn.py for training models and saving them using various datasets.

• mnist_models_testing.ipynb
  A notebook used for testing pre-trained models on the MNIST dataset.

• mlp_tf.py
  Tenserflow version

python version 3.9.x install required packages: pip install scikit-learn matplotlib numpy opencv-python

for running tensorflow version or training models (optional): pip install tensorflow

run python draw_digit to draw digits