Popular Courses
Beginner – Linear Regression
Introduction
Beginner – Decision Trees
Introduction
Beginner – Neural Networks
Introduction

Sign in to Curvingo

or

New Account     Reset Password     

Plans

Details
Testing Your AI
Introduction
Our Previous Approach
K-Fold Cross-Validation
Leave-One-Out Cross Validation (LOOCV)
See For Yourself
Which K Should You Pick?
Conclusion
Code with Python
Summary
Curvingo · Made in India

Share

Send us Feedback

Please let us know if you have any feedback and suggestions, or if you find any errors and bugs in our content.

Sorry, your message couldn’t be submitted. Please try again!

Thanks for your feedback!

Reset Progress

This will delete your progress and chat data for all chapters in this course, and cannot be undone!

Glossary

Select one of the keywords on the left…

Testing Your AIK-Fold Cross-Validation

Reading time: ~5 min

Instead of hoping for a fair split, let every farm take a turn in the validation set. That is k-fold cross-validation.

Here is the recipe:

  1. Cut the data into k equal parts, called folds.
  2. Hold out one fold for validation, train on all the others, and score the model on the held-out fold.
  3. Repeat with each fold in turn, starting a fresh model every time, until every fold has been held out once.
  4. Average the k scores. That average is the cross-validation score.

Each row below is one round: the teal squares are the farms held out, the blue ones the farms the model trains on. The chart under the button keeps the average error of every k you run. Choose k, run the folds and complete both missions.

Number of folds: k = ${k}

With 5 folds the five scores were 0.72, 0.30, 0.08, 0.34 and 0.48 — as wide a spread as the random splits. But their average, 0.38, uses every farm once for validation, so it does not depend on one lucky or unlucky fold. The catch: k-fold trains a new model , so it costs k times as much computing.

Sina