KwickClips Artificial Intelligence · 41 sec · free
Is a lower MSE better?
Short answer. Yes. Lower error is better.
What data should you test on?
Unseen data only.
MSE, RMSE and splits
| Errors 2, 2, 2, 2: MSE = 4 |
| RMSE = square root = 2 |
| Test on unseen data only |
Remember
| Lower error is better |
| Train, validate, test |
How do you score a model that predicts marks? Use MSE and RMSE. Square each error and average: MSE is four. Its square root, RMSE, is two. Always test on unseen data. Lower is better. Train, validate, then test.
This clip is from the full lesson: Model Evaluation Metrics: Confusion Matrix, Accuracy, Precision, Recall and F1 — 8 minutes, with the tables, the quick answers and the whole lesson in text.
Useful for: CBSE Class 10 Artificial Intelligence (417), CBSE Class 10 Artificial Intelligence (417), CBSE Class 10 Artificial Intelligence (417), CBSE Class 10 Artificial Intelligence (417)
More KwickClips from this lesson
Why is it called a confusion matrix?38 sec
What is the F1 formula?43 sec
Can a 99% accurate model be useless?42 secDisclaimer. KwickAcademy is free study material for general learning and revision. Parts of it, including the voice-over, are produced with the help of AI tools and may contain errors; if you spot one, please tell us and we will correct it. Syllabus, marks and exam details follow the latest official board publications available to us, and boards can change them at any time, so always confirm against your board's official website and your school. Using this material does not guarantee any marks or result. Board names and trademarks belong to their owners; Kwickprep is not affiliated with or endorsed by any examination board. We never ask for passwords, OTPs or ID numbers. Your progress is saved only in this browser. Full disclaimer · Privacy