Timeline for Showing the Equivalence Between the $ {L}_{2} $ Norm Regularized Regression and $ {L}_{2} $ Norm Constrained Regression Using KKT
Current License: CC BY-SA 4.0
8 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 8, 2019 at 15:31 | comment | added | stats_model | The KKT conditions in this case are a generalization of the “first order conditions” I mention by differentiating the Lagrangian and setting the derivative equal to 0. Since in this example, the constraints hold with equality, we don’t need the KKT conditions in full generally. In more complicated cases, all that happens is that some of the equalities above become inequalities and the multiplier becomes 0 for constraints become non binding . For example, this is exactly what happens when $M > ||\beta^{OLS}||$ in the above. | |
| Apr 8, 2019 at 12:28 | comment | added | jeza | many thanks, why you do not mention KKT? I am not familiar with this area, so treat me as a high school student. | |
| Apr 8, 2019 at 0:42 | history | edited | stats_model | CC BY-SA 4.0 |
added 213 characters in body
|
| Apr 8, 2019 at 0:31 | history | edited | stats_model | CC BY-SA 4.0 |
added 2186 characters in body
|
| Apr 7, 2019 at 21:41 | comment | added | jeza | could you please provide us with a detailed answer step by step with a practical example if that possible. | |
| Apr 4, 2019 at 18:42 | history | edited | stats_model | CC BY-SA 4.0 |
added 252 characters in body
|
| Apr 4, 2019 at 16:47 | history | edited | stats_model | CC BY-SA 4.0 |
added 5 characters in body
|
| Apr 4, 2019 at 16:34 | history | answered | stats_model | CC BY-SA 4.0 |