MOOC

  1. Deep Learning .ai, be prepared for a lot of math
  2. Fast AI, be ready to roll your sleeves and code in python, current is based on pytorch
  3. Google Machine Learning Crash Crouse
  4. Udacity - Intro to deep learning with pytorch
  5. Edx - Reinforcement Learning Explained
  6. AWS Machine Learning
  7. Kaggle Learn
  8. Reproducible Research (part of data science specialization)
  9. Developing Data Product (part of data science specialization)

Resources

  1. Tensorflow, harder to use when it is introduced, getting better
  2. PyTorch, very pythonic
  3. sk-learn, not a fan, but worth to mention
  4. xgboost, gbm

Topics

  1. data preparation, transformation, imputation, normalization
  2. model comparison, evaluation
  3. model interpretability
  4. productionization and reproducibility
  5. AutoML is a thing, worth look into

Recommendations

Write your own perceptron, loss function, optimizer, etc. for an plain vanilla network to get started.
And then get into imagine recognition, speech recognition, time series, RL with existing libraries.
And also be ready to read research papers, heck lotsa papers. :=)
Do an kaggle competition!