DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Yun, Se-Young | - |
dc.contributor.advisor | 윤세영 | - |
dc.contributor.author | Kim, Sungnyun | - |
dc.date.accessioned | 2022-04-13T05:40:07Z | - |
dc.date.available | 2022-04-13T05:40:07Z | - |
dc.date.issued | 2021 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=963739&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/292504 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : AI대학원, 2021.8,[iv, 31 p. :] | - |
dc.description.abstract | As numerous meta-learning algorithms improve performance when solving few-shot classification problems for practical applications, accurate prediction of uncertainty has been considered essential. In meta-training, the algorithm treats all the generated tasks equally and updates the model to perform well on the training tasks. For the training, some of the tasks might be hard for the model to infer the query examples from the support examples, especially when a huge mismatch between the support set and the query set exists. The distribution mismatch makes the model have wrong confidences that cause a calibration problem. In this study, we propose a novel meta-training method that measures the distribution mismatch and lets the model predict with more careful confidence. Moreover, our method is algorithm-agnostic and readily expanded to include a range of meta-learning models. Through extensive experiments, including dataset shift, we present that our training strategy helps the model avoid being indiscriminately confident, and thereby, produce calibrated classification results without the loss of accuracy. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | Few-Shot Learning▼aMeta-Learning▼aConfidence Calibration▼aDistribution Mismatch▼aTask Uncertainty | - |
dc.subject | 퓨샷 학습▼a메타 학습▼a확신도 교정▼a분포 불일치▼a태스크 불확실성 | - |
dc.title | Calibration of few-shot classification tasks | - |
dc.title.alternative | 퓨샷 분류 태스크 교정: 분포 불일치에 의한 확신도 오류 완화 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :AI대학원, | - |
dc.contributor.alternativeauthor | 김성년 | - |
dc.title.subtitle | mitigating misconfidence from distribution mismatch | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.