ETRI-Knowledge Sharing Plaform

ENGLISH

성과물

논문 검색
구분 SCI
연도 ~ 키워드

상세정보

학술대회 A Multi-Scaled Method for Parallel Bayesian Optimization in Deep Predictive Analytics
Cited 2 time in scopus Download 2 time Share share facebook twitter linkedin kakaostory
저자
문용혁, 이용주
발행일
202010
출처
International Conference on Information and Communication Technology Convergence (ICTC) 2020, pp.1306-1308
DOI
https://dx.doi.org/10.1109/ICTC49870.2020.9289573
협약과제
20HS3300, 부하분산과 능동적 적시 대응을 위한 빅데이터 엣지 분석 기술 개발, 이용주
초록
This paper proposes a new roll-out scaling method for parallel Bayesian optimization and discusses how the proposed multi-scaled optimization guarantees a better convergence speed with outperformed accuracy than the conventional parallel search algorithms. Experiment results demonstrate that an entire search space can be efficiently reduced to more feasible subdomains. The performance of parallel Bayesian search can be further accelerated based on the interchangeable local evidence by properly adjusting three quantitative aspects in terms of space factorization, search direction, and architecture scaling.
키워드
Bayesian optimization, black-box learning algorithm, deep learning, hyperparameters
KSP 제안 키워드
Bayesian Optimization, Bayesian search, Black Box, Experiment results, Multi-scaled, Parallel search, Scaling method, Search Algorithm(GSA), Search Space, Search direction, convergence speed