Ridge regression: Difference between revisions
From Rice Wiki
(Created page with "'''Ridge regression''' is a regression model with an additional term called the ''regularizer''. The motive for this model is to discourage overfitting. It is similar to Regularization, except it more heavily punishes complex models. = Regularizer = The '''regularizer''' <math>\lambda</math> is an additional term to the loss function that penalizes higher order terms. <math>\lambda \sum \left| w_j^2 \right|</math>") |
No edit summary |
||
| (2 intermediate revisions by the same user not shown) | |||
| Line 1: | Line 1: | ||
'''Ridge regression''' is a regression model with an additional term called the ''regularizer''. The motive for this model is to discourage [[Curve Fitting#Overfitting|overfitting]]. It is similar to [[ | '''Ridge regression''' is a regression model with an additional term called the ''regularizer''. The motive for this model is to discourage [[Curve Fitting#Overfitting|overfitting]]. It is similar to [[lasso regression]], except it more heavily punishes complex models. | ||
= Regularizer = | = Regularizer = | ||
| Line 5: | Line 5: | ||
<math>\lambda \sum \left| w_j^2 \right|</math> | <math>\lambda \sum \left| w_j^2 \right|</math> | ||
[[Category:Machine Learning]] | |||
Latest revision as of 19:59, 18 May 2024
Ridge regression is a regression model with an additional term called the regularizer. The motive for this model is to discourage overfitting. It is similar to lasso regression, except it more heavily punishes complex models.
Regularizer
The regularizer is an additional term to the loss function that penalizes higher order terms.
