Deep Learning Specialisation

Coursera Specialisation (Andrew Ng)

  • Using Python, NumPy, Tensorflow and Keras in Jupyter Notebooks
  • The honour code does not permit making solutions public so I cannot share assignments on GitHub
  • Course Notes

A series of 5 courses make up this specialisation

  1. Neural Networks & Deep Learning
    • Introduction
    • Basics
      • Assignment: Logistic Regression with a Neural Network Mindset
    • One hidden layer
      • Assignment: Planar Data Classification with One Hidden Layer
    • Deep Neural Networks
      • Assignment: Building a Neural Network- Step by Step
      • Assignment: Deep Neural Network for Image Classification
  2. Improving Deep Neural Networks: Hyperparameter Tuning, Regularisation & Optimisation
    • Practical aspects
      • Assignment: Initialisation
      • Assignment: Regularisation
      • Assignment: Gradient Checking
    • Optimisation algorithms
      • Assignment: Optimisation Methods
    • Hyperparameter tuning, batch normalisation & programming frameworks
      • Assignment: TensorFlow Tutorial
  3. Structuring Machine Learning Projects
    • Best practice
    • Differing distributions
    • Bias / variance analysis
    • Error analysis
    • Transfer learning & multi-task learning
    • End to end deep learning
  4. Convolutional Neural Networks
    • Foundations
      • Assignment: Building a Convolution Model- Step by Step
      • Assignment: Convolution Model for Image Classification
    • Deep convolutional model case studies
      • Assignment: Keras Tutorial – Detecting Happy Faces
      • Assignment: Residual Networks
    • Object detection
      • Assignment: Autonomous Driving – Car Detection
    • Special applications: Face recognition & neural style transfer
      • Assignment: Art Generation with Neural Style Transfer
  5. Sequence Models
    • RNNs (LSTM models)
      • Assignment: Building a Recurrent Neural Network – Step by Step
      • Assignment: Character level language model – Inventing Dinosaur Names
      • Assignment: Improvise a Jazz Solo with an LSTM Network
    • Natural language processing & word embeddings
      • Assignment: Operations on word vectors
      • Assignment: Emojis from Word Vector Representations
    • Sequence models & attention mechanism
      • Assignment: Neural Machine Translation with Attention
      • Assignment: Trigger Word Detection