Matches in SemOpenAlex for { <https://semopenalex.org/work/W4289639547> ?p ?o ?g. }
Showing items 1 to 86 of
86
with 100 items per page.
- W4289639547 endingPage "13" @default.
- W4289639547 startingPage "1" @default.
- W4289639547 abstract "Hyperparameter optimization (HPO), characterized by hyperparameter tuning, is not only a critical step for effective modeling but also is the most time-consuming process in machine learning. Traditional search-based algorithms tend to require extensive configuration evaluations for each round to select the desirable hyperparameters during the process, and they are often very inefficient for the implementations on large-scale tasks. In this paper, we study the HPO problem via meta-learning (MtL) approach under the low-rank tensor completion (LRTC) framework. Our proposed approach predicts the performance for hyperparameters of new problems based on their previous performance so that the underlying suitable hyperparameters with better efficiency can be attained. Different from existing approaches, the hyperparameter performance space is instantiated under tensor framework that can preserve the spatial structure and reflect the correlations among the adjacent hyperparameters. When some partial evaluations are available for a new problem, the task of estimating the performance of the unevaluated hyperparameters can be formulated as a tensor completion (TC) problem. Toward the completion purpose, we develop an LRTC algorithm utilizing the sum of nuclear norm (SNN) model. A kernelized version is further developed to capture the nonlinear structure of the performance space. In addition, a corresponding coupled matrix factorization (CMF) algorithm is established to render the predictions solely depend on the meta-features to avoid additional hyperparameter evaluations. Finally, a strategy integrating LRTC and CMF is provided to further enhance the recommendation capacity. We test recommendation performance with our proposed methods for classical SVM and the state-of-the-art deep neural networks such as vision transformer (ViT) and residual network (ResNet), and the obtained results demonstrate the effectiveness of our approaches under various evaluation metrics by comparing with the baselines commonly used for MtL." @default.
- W4289639547 created "2022-08-03" @default.
- W4289639547 creator A5006353150 @default.
- W4289639547 creator A5008007155 @default.
- W4289639547 date "2022-01-01" @default.
- W4289639547 modified "2023-10-16" @default.
- W4289639547 title "A New Automatic Hyperparameter Recommendation Approach Under Low-Rank Tensor Completion Framework" @default.
- W4289639547 cites W1492459858 @default.
- W4289639547 cites W1495775210 @default.
- W4289639547 cites W1507030697 @default.
- W4289639547 cites W1512022526 @default.
- W4289639547 cites W1558866804 @default.
- W4289639547 cites W1692958259 @default.
- W4289639547 cites W1997201895 @default.
- W4289639547 cites W2010770252 @default.
- W4289639547 cites W2078677240 @default.
- W4289639547 cites W2089213632 @default.
- W4289639547 cites W2091449379 @default.
- W4289639547 cites W2112204321 @default.
- W4289639547 cites W2122379760 @default.
- W4289639547 cites W2157069634 @default.
- W4289639547 cites W2157686535 @default.
- W4289639547 cites W2194775991 @default.
- W4289639547 cites W2344468249 @default.
- W4289639547 cites W2563364594 @default.
- W4289639547 cites W2762772695 @default.
- W4289639547 cites W2950220059 @default.
- W4289639547 cites W2964214749 @default.
- W4289639547 cites W2982471955 @default.
- W4289639547 cites W3000276536 @default.
- W4289639547 cites W3204801262 @default.
- W4289639547 cites W4211208325 @default.
- W4289639547 cites W4312769570 @default.
- W4289639547 doi "https://doi.org/10.1109/tpami.2022.3195658" @default.
- W4289639547 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/35914040" @default.
- W4289639547 hasPublicationYear "2022" @default.
- W4289639547 type Work @default.
- W4289639547 citedByCount "0" @default.
- W4289639547 crossrefType "journal-article" @default.
- W4289639547 hasAuthorship W4289639547A5006353150 @default.
- W4289639547 hasAuthorship W4289639547A5008007155 @default.
- W4289639547 hasConcept C10485038 @default.
- W4289639547 hasConcept C119857082 @default.
- W4289639547 hasConcept C121332964 @default.
- W4289639547 hasConcept C12267149 @default.
- W4289639547 hasConcept C154945302 @default.
- W4289639547 hasConcept C155281189 @default.
- W4289639547 hasConcept C163716315 @default.
- W4289639547 hasConcept C202444582 @default.
- W4289639547 hasConcept C2778459887 @default.
- W4289639547 hasConcept C33923547 @default.
- W4289639547 hasConcept C41008148 @default.
- W4289639547 hasConcept C62520636 @default.
- W4289639547 hasConcept C8642999 @default.
- W4289639547 hasConceptScore W4289639547C10485038 @default.
- W4289639547 hasConceptScore W4289639547C119857082 @default.
- W4289639547 hasConceptScore W4289639547C121332964 @default.
- W4289639547 hasConceptScore W4289639547C12267149 @default.
- W4289639547 hasConceptScore W4289639547C154945302 @default.
- W4289639547 hasConceptScore W4289639547C155281189 @default.
- W4289639547 hasConceptScore W4289639547C163716315 @default.
- W4289639547 hasConceptScore W4289639547C202444582 @default.
- W4289639547 hasConceptScore W4289639547C2778459887 @default.
- W4289639547 hasConceptScore W4289639547C33923547 @default.
- W4289639547 hasConceptScore W4289639547C41008148 @default.
- W4289639547 hasConceptScore W4289639547C62520636 @default.
- W4289639547 hasConceptScore W4289639547C8642999 @default.
- W4289639547 hasLocation W42896395471 @default.
- W4289639547 hasLocation W42896395472 @default.
- W4289639547 hasOpenAccess W4289639547 @default.
- W4289639547 hasPrimaryLocation W42896395471 @default.
- W4289639547 hasRelatedWork W2977967020 @default.
- W4289639547 hasRelatedWork W3033561277 @default.
- W4289639547 hasRelatedWork W3141246024 @default.
- W4289639547 hasRelatedWork W3186782765 @default.
- W4289639547 hasRelatedWork W4280535922 @default.
- W4289639547 hasRelatedWork W4283697347 @default.
- W4289639547 hasRelatedWork W4287514708 @default.
- W4289639547 hasRelatedWork W4295309597 @default.
- W4289639547 hasRelatedWork W4298144215 @default.
- W4289639547 hasRelatedWork W4320494184 @default.
- W4289639547 isParatext "false" @default.
- W4289639547 isRetracted "false" @default.
- W4289639547 workType "article" @default.