Deep Learning using Keras in Python

Constantly updated repository

Get a basic overview of Python’s Keras library to implement ANN, CNN and RNN models. Source: Keras website, plus additional used libraries github source.
Ipython notebooks:

• Notes on Perceptron, MLP, CNN and RNN
• Rosenblatt Perceptron
• T-SNE embedding visualization on Fashion MNIST
• Comparison of various optmization algorithms
• Multi-Layer Perceptron classification on Fashion MNIST
• MLP hyper-paramter optimization
  • Initializing and maintaining learning rate
  • Initializing biases
  • Initializing weights
  • L1, L2, L1L2 Regularizations
  • Dropout regularization
  • Early stopping
• CNN on CIFAR10 with visualizations