/Digit-Recognizer

Digit Recognition on MNIST Data

Primary LanguageJupyter Notebook

Digit-Recognizer

Digit Recognition on MNIST Data.
If you want to use the trained weights for your own testing, initialize all the variables first and then restore the saved model. It will overwrite the pre-weights and you will have the trained weights for your own testing purpose.

Libraries Used:

  • Numpy
  • Tensorflow

Network:

  • Multilayer Perceptron

Optimizer:

  • AdamOptimizer

Hidden Layers:

  • 2

No. of nodes in both Hidden Layers:

  • 256