Statements (48)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
Machine Learning Algorithm
Regression Method Statistical Technique |
| gptkbp:alsoKnownAs |
Tikhonov Regularization
|
| gptkbp:appliesTo |
gptkb:Linear_Regression
|
| gptkbp:category |
Supervised Learning
Linear Models Regularization Methods |
| gptkbp:contrastsWith |
gptkb:Elastic_Net
gptkb:Ordinary_Least_Squares gptkb:Lasso_Regression |
| gptkbp:feature |
Does Not Set Coefficients Exactly to Zero
Handles Many Predictors Improves Prediction Accuracy Sensitive to Scaling of Predictors Shrinks Coefficient Estimates |
| gptkbp:form |
(X^T X + λI)^{-1} X^T y
|
| https://www.w3.org/2000/01/rdf-schema#label |
Ridge Regression
|
| gptkbp:implementedIn |
gptkb:SAS
gptkb:MATLAB gptkb:Stata gptkb:scikit-learn R |
| gptkbp:improves |
Bias of Estimates
|
| gptkbp:introduced |
gptkb:Andrey_Tikhonov
|
| gptkbp:introducedIn |
1943
|
| gptkbp:parameterSelectionMethod |
Cross-Validation
|
| gptkbp:penaltyTerm |
Lambda times Sum of Squared Coefficients
|
| gptkbp:reduces |
Sum of Squared Residuals plus L2 Penalty
Variance of Estimates |
| gptkbp:regularizationType |
L2 Regularization
|
| gptkbp:relatedTo |
gptkb:Elastic_Net
gptkb:Ordinary_Least_Squares gptkb:Lasso_Regression Principal Component Regression |
| gptkbp:requires |
Selection of Regularization Parameter
|
| gptkbp:solvedBy |
Analytical Solution
Closed Form Solution |
| gptkbp:usedFor |
Regression Analysis
Overfitting Reduction Multicollinearity Handling |
| gptkbp:usedIn |
gptkb:Machine_Learning
Finance Statistics Genomics Econometrics |
| gptkbp:bfsParent |
gptkb:Linear_Model
|
| gptkbp:bfsLayer |
5
|