Matches in SemOpenAlex for { <https://semopenalex.org/work/W2804162248> ?p ?o ?g. }
- W2804162248 abstract "Cross-validation is the workhorse of modern applied statistics and machine learning, as it provides a principled framework for selecting the model that maximizes generalization performance. In this paper, we show that the cross-validation risk is differentiable with respect to the hyperparameters and training data for many common machine learning algorithms, including logistic regression, elastic-net regression, and support vector machines. Leveraging this property of differentiability, we propose a cross-validation gradient method (CVGM) for hyperparameter optimization. Our method enables efficient optimization in high-dimensional hyperparameter spaces of the cross-validation risk, the best surrogate of the true generalization ability of our learning algorithm." @default.
- W2804162248 created "2018-06-01" @default.
- W2804162248 creator A5023055112 @default.
- W2804162248 creator A5059803135 @default.
- W2804162248 date "2018-05-18" @default.
- W2804162248 modified "2023-09-27" @default.
- W2804162248 title "Optimizing for Generalization in Machine Learning with Cross-Validation Gradients." @default.
- W2804162248 cites W1480376833 @default.
- W2804162248 cites W1529817821 @default.
- W2804162248 cites W1868018859 @default.
- W2804162248 cites W2035042171 @default.
- W2804162248 cites W2101234009 @default.
- W2804162248 cites W2122825543 @default.
- W2804162248 cites W2124022122 @default.
- W2804162248 cites W2139338362 @default.
- W2804162248 cites W2158001550 @default.
- W2804162248 cites W2158915909 @default.
- W2804162248 cites W2187061624 @default.
- W2804162248 cites W2592457170 @default.
- W2804162248 cites W2796676459 @default.
- W2804162248 cites W2950182411 @default.
- W2804162248 hasPublicationYear "2018" @default.
- W2804162248 type Work @default.
- W2804162248 sameAs 2804162248 @default.
- W2804162248 citedByCount "0" @default.
- W2804162248 crossrefType "posted-content" @default.
- W2804162248 hasAuthorship W2804162248A5023055112 @default.
- W2804162248 hasAuthorship W2804162248A5059803135 @default.
- W2804162248 hasConcept C10485038 @default.
- W2804162248 hasConcept C105795698 @default.
- W2804162248 hasConcept C111472728 @default.
- W2804162248 hasConcept C115903097 @default.
- W2804162248 hasConcept C119857082 @default.
- W2804162248 hasConcept C12267149 @default.
- W2804162248 hasConcept C127413603 @default.
- W2804162248 hasConcept C134306372 @default.
- W2804162248 hasConcept C138885662 @default.
- W2804162248 hasConcept C14948415 @default.
- W2804162248 hasConcept C151956035 @default.
- W2804162248 hasConcept C154945302 @default.
- W2804162248 hasConcept C177148314 @default.
- W2804162248 hasConcept C189950617 @default.
- W2804162248 hasConcept C201995342 @default.
- W2804162248 hasConcept C202615002 @default.
- W2804162248 hasConcept C27181475 @default.
- W2804162248 hasConcept C2780451532 @default.
- W2804162248 hasConcept C2781002164 @default.
- W2804162248 hasConcept C33923547 @default.
- W2804162248 hasConcept C41008148 @default.
- W2804162248 hasConcept C77967617 @default.
- W2804162248 hasConcept C83546350 @default.
- W2804162248 hasConcept C8642999 @default.
- W2804162248 hasConceptScore W2804162248C10485038 @default.
- W2804162248 hasConceptScore W2804162248C105795698 @default.
- W2804162248 hasConceptScore W2804162248C111472728 @default.
- W2804162248 hasConceptScore W2804162248C115903097 @default.
- W2804162248 hasConceptScore W2804162248C119857082 @default.
- W2804162248 hasConceptScore W2804162248C12267149 @default.
- W2804162248 hasConceptScore W2804162248C127413603 @default.
- W2804162248 hasConceptScore W2804162248C134306372 @default.
- W2804162248 hasConceptScore W2804162248C138885662 @default.
- W2804162248 hasConceptScore W2804162248C14948415 @default.
- W2804162248 hasConceptScore W2804162248C151956035 @default.
- W2804162248 hasConceptScore W2804162248C154945302 @default.
- W2804162248 hasConceptScore W2804162248C177148314 @default.
- W2804162248 hasConceptScore W2804162248C189950617 @default.
- W2804162248 hasConceptScore W2804162248C201995342 @default.
- W2804162248 hasConceptScore W2804162248C202615002 @default.
- W2804162248 hasConceptScore W2804162248C27181475 @default.
- W2804162248 hasConceptScore W2804162248C2780451532 @default.
- W2804162248 hasConceptScore W2804162248C2781002164 @default.
- W2804162248 hasConceptScore W2804162248C33923547 @default.
- W2804162248 hasConceptScore W2804162248C41008148 @default.
- W2804162248 hasConceptScore W2804162248C77967617 @default.
- W2804162248 hasConceptScore W2804162248C83546350 @default.
- W2804162248 hasConceptScore W2804162248C8642999 @default.
- W2804162248 hasLocation W28041622481 @default.
- W2804162248 hasOpenAccess W2804162248 @default.
- W2804162248 hasPrimaryLocation W28041622481 @default.
- W2804162248 hasRelatedWork W1563195889 @default.
- W2804162248 hasRelatedWork W1971788608 @default.
- W2804162248 hasRelatedWork W1989048657 @default.
- W2804162248 hasRelatedWork W2006151125 @default.
- W2804162248 hasRelatedWork W2067467251 @default.
- W2804162248 hasRelatedWork W2107261944 @default.
- W2804162248 hasRelatedWork W2368980208 @default.
- W2804162248 hasRelatedWork W2399786745 @default.
- W2804162248 hasRelatedWork W2411613456 @default.
- W2804162248 hasRelatedWork W2524816779 @default.
- W2804162248 hasRelatedWork W2780562800 @default.
- W2804162248 hasRelatedWork W2782034465 @default.
- W2804162248 hasRelatedWork W2798705318 @default.
- W2804162248 hasRelatedWork W2884898416 @default.
- W2804162248 hasRelatedWork W2890035728 @default.
- W2804162248 hasRelatedWork W2890787084 @default.
- W2804162248 hasRelatedWork W2964328735 @default.
- W2804162248 hasRelatedWork W3014048727 @default.
- W2804162248 hasRelatedWork W3165644918 @default.
- W2804162248 hasRelatedWork W3212292802 @default.
- W2804162248 isParatext "false" @default.