This lesson is being piloted (Beta version)

C3DIS 2019 - Introduction to Machine Learning: Glossary

Key Points

Introduction
  • ML algorithms learn from data instead of being human-programmed

  • A large amount of quality data is essential for ML

Machine Learning Overview
  • There are many elements to a ML project, the training and algorithms are a very small part of that.

Machine Learning Metrics for Performance
  • Lot’s of different metrics, which one you use ultimately depends on the application

  • Variance / bias trade off

Machine Learning Metrics for Performance
  • Lot’s of different metrics, which one you use ultimately depends on the application

  • Variance / bias trade off

Machine Learning Model Selection and Validation
  • Traditional statistical methods for validation are still extremely useful.

  • Most of the time your data is much more important that the model itself.

Glossary

FIXME