Machine Learning Theory - Underfitting vs Overfitting

preview_player
Показать описание
What is Underfitting and Overfitting? | machine learning explained | overfitting machine learning Illustrated using Lego pieces and diagrams.

What is Underfitting?
Oversimplifying the problem
Does not do well in the training set
Error due to bias

What is Overfitting?
High Variance complicates the problem more than necessary to perform well on the training set.
Рекомендации по теме
Комментарии
Автор

Best explanation for overfitting and underfitting on YouTube.

stanlukash
Автор

Excellent explanation and examples. Thank you so much!

bryanirvine
Автор

Great job! Well explained with the pictures. Kudos

rafsangoni
Автор

What method can you use to fix underfitting in r?

battles
Автор

When simplifying or complicating, why is the fitting excessive?
The explanation started with the filter "LEGO", and the LEGO Figure was misplaced into the "Not LEGO" category. We say it underfits.
It became "LEGO Brick", now the figure is correctly in "Not LEGO". It, we say, loosely fits.
Then it became "LEGO Brick with four studs", the figure is still in the correct category, but bricks with more than 4 studs (or none) and literally everything else that isn't a 4-stud brick is excluded. We say this overfits.
I'm asking why this is and should be considered excessive/unnecessary. Why under or over?
The way I see it, there's a motivated invisible hand external to the model who believes it's "correct" or "incorrect" according to what they want out of the model, which is to include LEGO Bricks with at least or more than four studs, and not Figures.

TheAlison
Автор

thank you for the good explanation video

oompster