FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2070598047> ?p ?o ?g. }

• W2070598047 abstract "Many complex and high dimensional real-world classification problems require a carefully chosen set of features,algorithms and hyperparameters to achieve the desired generalization performance. The choice of asuitable feature representation has a great effect on the prediction performance. Manifold learning techniquesâ?? like PCA, Isomap, Local Linear Embedding (LLE) or Autoencoders â?? are able to learn a better suitablerepresentation automatically. However, the performance of a manifold learner heavily depends on the dataset.This paper presents a novel automatic optimization framework that incorporates multiple manifold learningalgorithms in a holistic classification pipeline together with feature selection and multiple classifiers with arbitraryhyperparameters. The highly combinatorial optimization problem is solved efficiently using evolutionaryalgorithms. Additionally, a multi-pipeline classifier based on the optimization trajectory is presented. Theevaluation on several datasets shows that the proposed framework outperforms the Auto-WEKA frameworkin terms of generalization and optimization speed in many cases." @default.
• W2070598047 created "2016-06-24" @default.
• W2070598047 creator A5037031302 @default.
• W2070598047 creator A5050558405 @default.
• W2070598047 date "2015-01-01" @default.
• W2070598047 modified "2023-09-26" @default.
• W2070598047 title "Representation Optimization with Feature Selection and Manifold Learning in a Holistic Classification Framework" @default.
• W2070598047 cites W1535654799 @default.
• W2070598047 cites W1576660662 @default.
• W2070598047 cites W1663973292 @default.
• W2070598047 cites W1750919668 @default.
• W2070598047 cites W1919621926 @default.
• W2070598047 cites W1966681180 @default.
• W2070598047 cites W2001141328 @default.
• W2070598047 cites W2001619934 @default.
• W2070598047 cites W2020355555 @default.
• W2070598047 cites W2041785029 @default.
• W2070598047 cites W2097998348 @default.
• W2070598047 cites W2100483895 @default.
• W2070598047 cites W2100495367 @default.
• W2070598047 cites W2102017903 @default.
• W2070598047 cites W2102539288 @default.
• W2070598047 cites W2103250033 @default.
• W2070598047 cites W2104752854 @default.
• W2070598047 cites W2106053110 @default.
• W2070598047 cites W2106411961 @default.
• W2070598047 cites W2111072639 @default.
• W2070598047 cites W2112148214 @default.
• W2070598047 cites W2132549764 @default.
• W2070598047 cites W2133361079 @default.
• W2070598047 cites W2133487567 @default.
• W2070598047 cites W2137570937 @default.
• W2070598047 cites W2140095548 @default.
• W2070598047 cites W2144935315 @default.
• W2070598047 cites W2149551942 @default.
• W2070598047 cites W2153934661 @default.
• W2070598047 cites W2154872931 @default.
• W2070598047 cites W2156718197 @default.
• W2070598047 cites W2156838815 @default.
• W2070598047 cites W2157801062 @default.
• W2070598047 cites W2163922914 @default.
• W2070598047 cites W2166107799 @default.
• W2070598047 cites W2294798173 @default.
• W2070598047 cites W3120740533 @default.
• W2070598047 cites W3261618 @default.
• W2070598047 cites W434012021 @default.
• W2070598047 doi "https://doi.org/10.5220/0005183600350044" @default.
• W2070598047 hasPublicationYear "2015" @default.
• W2070598047 type Work @default.
• W2070598047 sameAs 2070598047 @default.
• W2070598047 citedByCount "4" @default.
• W2070598047 countsByYear W20705980472015 @default.
• W2070598047 countsByYear W20705980472016 @default.
• W2070598047 countsByYear W20705980472018 @default.
• W2070598047 crossrefType "proceedings-article" @default.
• W2070598047 hasAuthorship W2070598047A5037031302 @default.
• W2070598047 hasAuthorship W2070598047A5050558405 @default.
• W2070598047 hasConcept C11413529 @default.
• W2070598047 hasConcept C119857082 @default.
• W2070598047 hasConcept C134306372 @default.
• W2070598047 hasConcept C137836250 @default.
• W2070598047 hasConcept C139532973 @default.
• W2070598047 hasConcept C148483581 @default.
• W2070598047 hasConcept C151876577 @default.
• W2070598047 hasConcept C153180895 @default.
• W2070598047 hasConcept C154945302 @default.
• W2070598047 hasConcept C177148314 @default.
• W2070598047 hasConcept C2778626561 @default.
• W2070598047 hasConcept C33923547 @default.
• W2070598047 hasConcept C41008148 @default.
• W2070598047 hasConcept C41608201 @default.
• W2070598047 hasConcept C59404180 @default.
• W2070598047 hasConcept C70518039 @default.
• W2070598047 hasConcept C8642999 @default.
• W2070598047 hasConcept C95623464 @default.
• W2070598047 hasConceptScore W2070598047C11413529 @default.
• W2070598047 hasConceptScore W2070598047C119857082 @default.
• W2070598047 hasConceptScore W2070598047C134306372 @default.
• W2070598047 hasConceptScore W2070598047C137836250 @default.
• W2070598047 hasConceptScore W2070598047C139532973 @default.
• W2070598047 hasConceptScore W2070598047C148483581 @default.
• W2070598047 hasConceptScore W2070598047C151876577 @default.
• W2070598047 hasConceptScore W2070598047C153180895 @default.
• W2070598047 hasConceptScore W2070598047C154945302 @default.
• W2070598047 hasConceptScore W2070598047C177148314 @default.
• W2070598047 hasConceptScore W2070598047C2778626561 @default.
• W2070598047 hasConceptScore W2070598047C33923547 @default.
• W2070598047 hasConceptScore W2070598047C41008148 @default.
• W2070598047 hasConceptScore W2070598047C41608201 @default.
• W2070598047 hasConceptScore W2070598047C59404180 @default.
• W2070598047 hasConceptScore W2070598047C70518039 @default.
• W2070598047 hasConceptScore W2070598047C8642999 @default.
• W2070598047 hasConceptScore W2070598047C95623464 @default.
• W2070598047 hasLocation W20705980471 @default.
• W2070598047 hasOpenAccess W2070598047 @default.
• W2070598047 hasPrimaryLocation W20705980471 @default.
• W2070598047 hasRelatedWork W1970769764 @default.
• W2070598047 hasRelatedWork W2013966960 @default.
• W2070598047 hasRelatedWork W2048701710 @default.
• W2070598047 hasRelatedWork W2062393208 @default.

• next