Model Evaluation#

How can we assess the quality of a machine learning model?
How can we detect underfitting or overfitting?
How can we adjust the algorithm to achieve a good fit?

This section will introduce key concepts and tools to evaluate and improve your model.

Learning outcomes:

  • Understand the purpose of splitting data into different sets (training, validation, test)

  • Become familiar with cross-validation techniques

  • Learn common metrics for regression and classification

  • Understand how the ROC curve is constructed step by step

  • Grasp the concepts of bias and variance and the tradeoff between them

  • Define overfitting and underfitting and identify when they occur

  • Know strategies to cope with high bias or high variance situations