10-601 Evaluation

From Cohen Courses
Jump to navigationJump to search

This a lecture used in the Syllabus for Machine Learning 10-601 in Fall 2014



  • Mitchell, Chapter 5.

What You Should Know Afterward

  • The difference between sample error and true error.
  • What a confidence interval is.
  • How to compute a confidence interval on the error rate of a classifier using the normal approximation.
  • What a paired test is, and how to compute use a paired test to compare two classifiers.
  • How to test the error rate of a classifier by cross-validation, or compare the error rates of two classifiers by cross-validation.