ml-intro

An introduction to machine learning

Resources

A Whirlwind Tour of Python by Jake VanderPlas

Python Data Science Handbook by Jake VanderPlas

Session 1

• Introduction to machine learning following notebooks/introduction.ipynb
• Supervised learning workflow following notebooks/heart-disease-linear.ipynb

Supervised learning workflow

1. Download the data
2. Prepare the data
   1. A look at the data files
   2. Loading the data with Pandas
   3. Getting an overview by visualizing the data
   4. Create feature matrix and target vector
   5. Create training and validation sets
3. Machine learning: Supervised classification
   1. Train a model
   2. Compare training and validation accuracy
   3. The confusion matrix
   4. Summarize the different scores

Session 2

• Inspect workflow steps, particularly loading, preparing and visualizing data
• Over- and underfitting Python Data Science Handbook 05.03-Hyperparameters-and-Model-Validation
• Support vector classifier Python Data Science Handbook 05.07-Support-Vector-Machines
• Random forest classifier Python Data Science Handbook 05.08-Random-Forests

Session 3

• Review of concepts
• Introduction to the lab
• Deep learning and introduction to medical imaging