/TensorFlow-deep-NN

A high-level tutorial into Deep Learning using MNIST data and TensorFlow library.

Primary LanguageJupyter Notebook

TensorFlow-deep-NN

A high-level tutorial into Deep Learning using MNIST data and TensorFlow library

Prerequisites :

  • Fundamental coding skills - learn
  • Linear algebra, especially matrix operations
  • Understanding how images are stored in computer memory.

Feel free to fork and adjust CONSTANTS to tweak network behaviour and explore how it changes algorithm performance and accuracy. Besides TensorFlow graph section can also be modified for learning purposes. It is highly recommended printing every variable that isn’t 100% clear for you.

Also, tensorboard can be used on a local environment for visualisation and debugging.