/NN_binary-classifier

NN-based binary classification with logistic regression loss, back-propagation with gradient descent and optional regularization

Primary LanguageJupyter NotebookOtherNOASSERTION

Contents

Python code implementation of a simple binary classification:

  • randomly generated data set
  • hyperbolic tangent (tanh) activation function (adequate for binary classification)
  • log or cross-entropy loss
  • NN - forward and back-propagation with gradient descent and optional regularization