FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2890061498> ?p ?o ?g. }

• W2890061498 abstract "In this dissertation, we focus on several important problems in structured prediction. In structured prediction, the label has a rich intrinsic substructure, and the loss varies with respect to the predicted label and the true label pair. Structured SVM is an extension of binary SVM to adapt to such structured tasks. In the first part of the dissertation, we study the surrogate losses and its efficient methods. To minimize the empirical risk, a surrogate loss which upper bounds the loss, is used as a proxy to minimize the actual loss. Since the objective function is written in terms of the surrogate loss, the choice of the surrogate loss is important, and the performance depends on it. Another issue regarding the surrogate loss is the efficiency of the argmax label inference for the surrogate loss. Efficient inference is necessary for the optimization since it is often the most time-consuming step. We present a new class of surrogate losses named bi-criteria surrogate loss, which is a generalization of the popular surrogate losses. We first investigate an efficient method for a slack rescaling formulation as a starting point utilizing decomposability of the model. Then, we extend the algorithm to the bi-criteria surrogate loss, which is very efficient and also shows performance improvements. In the second part of the dissertation, another important issue of regularization is studied. Specifically, we investigate a problem of regularization in hierarchical classification when a structural imbalance exists in the label structure. We present a method to normalize the structure, as well as a new norm, namely shared Frobenius norm. It is suitable for hierarchical classification that adapts to the data in addition to the label structure." @default.
• W2890061498 created "2018-09-27" @default.
• W2890061498 creator A5081997471 @default.
• W2890061498 date "2018-09-14" @default.
• W2890061498 modified "2023-09-27" @default.
• W2890061498 title "Efficient Structured Surrogate Loss and Regularization in Structured Prediction." @default.
• W2890061498 cites W1492459858 @default.
• W2890061498 cites W1515020792 @default.
• W2890061498 cites W1523949738 @default.
• W2890061498 cites W1574901103 @default.
• W2890061498 cites W1632114991 @default.
• W2890061498 cites W167016754 @default.
• W2890061498 cites W1768675166 @default.
• W2890061498 cites W1792316426 @default.
• W2890061498 cites W1929113426 @default.
• W2890061498 cites W1990061958 @default.
• W2890061498 cites W2004763266 @default.
• W2890061498 cites W2009570821 @default.
• W2890061498 cites W2014566476 @default.
• W2890061498 cites W2031248101 @default.
• W2890061498 cites W2040672759 @default.
• W2890061498 cites W2092423930 @default.
• W2890061498 cites W2102492119 @default.
• W2890061498 cites W2105644991 @default.
• W2890061498 cites W2111753845 @default.
• W2890061498 cites W2125993116 @default.
• W2890061498 cites W2127350043 @default.
• W2890061498 cites W2133590167 @default.
• W2890061498 cites W2135140174 @default.
• W2890061498 cites W2136064009 @default.
• W2890061498 cites W2142623206 @default.
• W2890061498 cites W2144578941 @default.
• W2890061498 cites W2150102617 @default.
• W2890061498 cites W2169819436 @default.
• W2890061498 cites W2301095666 @default.
• W2890061498 cites W2333790164 @default.
• W2890061498 cites W2508577564 @default.
• W2890061498 cites W2950080435 @default.
• W2890061498 cites W2952020226 @default.
• W2890061498 cites W2962805661 @default.
• W2890061498 cites W2963174605 @default.
• W2890061498 cites W2964205912 @default.
• W2890061498 cites W2964262905 @default.
• W2890061498 hasPublicationYear "2018" @default.
• W2890061498 type Work @default.
• W2890061498 sameAs 2890061498 @default.
• W2890061498 citedByCount "0" @default.
• W2890061498 crossrefType "posted-content" @default.
• W2890061498 hasAuthorship W2890061498A5081997471 @default.
• W2890061498 hasConcept C11413529 @default.
• W2890061498 hasConcept C119857082 @default.
• W2890061498 hasConcept C121332964 @default.
• W2890061498 hasConcept C126255220 @default.
• W2890061498 hasConcept C126838900 @default.
• W2890061498 hasConcept C131675550 @default.
• W2890061498 hasConcept C141341695 @default.
• W2890061498 hasConcept C142806159 @default.
• W2890061498 hasConcept C154945302 @default.
• W2890061498 hasConcept C158622935 @default.
• W2890061498 hasConcept C2776135515 @default.
• W2890061498 hasConcept C2776214188 @default.
• W2890061498 hasConcept C33923547 @default.
• W2890061498 hasConcept C41008148 @default.
• W2890061498 hasConcept C62520636 @default.
• W2890061498 hasConcept C71924100 @default.
• W2890061498 hasConceptScore W2890061498C11413529 @default.
• W2890061498 hasConceptScore W2890061498C119857082 @default.
• W2890061498 hasConceptScore W2890061498C121332964 @default.
• W2890061498 hasConceptScore W2890061498C126255220 @default.
• W2890061498 hasConceptScore W2890061498C126838900 @default.
• W2890061498 hasConceptScore W2890061498C131675550 @default.
• W2890061498 hasConceptScore W2890061498C141341695 @default.
• W2890061498 hasConceptScore W2890061498C142806159 @default.
• W2890061498 hasConceptScore W2890061498C154945302 @default.
• W2890061498 hasConceptScore W2890061498C158622935 @default.
• W2890061498 hasConceptScore W2890061498C2776135515 @default.
• W2890061498 hasConceptScore W2890061498C2776214188 @default.
• W2890061498 hasConceptScore W2890061498C33923547 @default.
• W2890061498 hasConceptScore W2890061498C41008148 @default.
• W2890061498 hasConceptScore W2890061498C62520636 @default.
• W2890061498 hasConceptScore W2890061498C71924100 @default.
• W2890061498 hasLocation W28900614981 @default.
• W2890061498 hasOpenAccess W2890061498 @default.
• W2890061498 hasPrimaryLocation W28900614981 @default.
• W2890061498 hasRelatedWork W1995572263 @default.
• W2890061498 hasRelatedWork W2070384470 @default.
• W2890061498 hasRelatedWork W2080905836 @default.
• W2890061498 hasRelatedWork W2180816288 @default.
• W2890061498 hasRelatedWork W2308398441 @default.
• W2890061498 hasRelatedWork W2344245780 @default.
• W2890061498 hasRelatedWork W2384341180 @default.
• W2890061498 hasRelatedWork W2677392994 @default.
• W2890061498 hasRelatedWork W2766674581 @default.
• W2890061498 hasRelatedWork W2912542961 @default.
• W2890061498 hasRelatedWork W2912983861 @default.
• W2890061498 hasRelatedWork W2962712569 @default.
• W2890061498 hasRelatedWork W3035812139 @default.
• W2890061498 hasRelatedWork W3116782167 @default.
• W2890061498 hasRelatedWork W3117809744 @default.
• W2890061498 hasRelatedWork W3119062064 @default.

• next