A paralleled embedding high-dimensional Bayesian optimization with additive Gaussian kernels for solving CNOP

doi:10.1016/j.ocemod.2023.102213

IOCAS-IR > 海洋环流与波动重点实验室

	A paralleled embedding high-dimensional Bayesian optimization with additive Gaussian kernels for solving CNOP
	Yuan, Shijin 1; Liu, Yaxuan 1; Qin, Bo 1; Mu, Bin 1; Zhang, Kun2,3
	2023-08-01
发表期刊	OCEAN MODELLING
ISSN	1463-5003
卷号	184 页码:26
通讯作者	Mu, Bin([email protected])
摘要	Conditional Nonlinear Optimal Perturbation (CNOP) is widely used in atmospheric and oceanic predictability studies. Solving CNOP is essentially a nonlinear optimization problem with certain constraints. One method to solve CNOP is the intelligent optimization algorithm. But it may not always obtain the satisfactory solution and efficiency because of the local exploitation of individual particles. The Bayesian optimization algorithm can avoid this local stagnation by building a global probability grasp for the optimization space of CNOP. Nonetheless, the limit of approximately 20-dimensional optimization space hampers its application in solving CNOP of high-dimensional numerical models. To overcome this bottleneck, we propose a paralleled embedding high-dimensional Bayesian optimization with additive Gaussian kernels (PEBOA) algorithm, which mainly consists of the feature extraction process and low-dimensional optimization process. In PEBOA, a feature extraction method (deepFE) using the convolutional Autoencoder with residual connections and customized constraint-preserved loss function is designed, which compresses 10 million dimensional data to relatively low -dimensional search spaces, from tens to hundreds of dimensions. Then, we propose the strategy of additive Gaussian kernels, with which the Bayesian optimization algorithm can optimize in relatively low-dimensional search spaces. Concretely, a kernel handles a subspace, and a subspace can handle dimensions approximately not more than 20. Additionally, the modified acquisition function can sample multiple candidates simulta-neously with the aim of accelerating the optimization process. We apply PEBOA to solve CNOP of Regional Ocean Modeling System (ROMS) for identifying optimal initial errors of upstream Kuroshio transport variation, which is a 10 million dimensional and time-consuming problem. The computational performance of PEBOA and physical mechanisms of the obtained CNOP are analyzed. Experimental results indicate that deepFE excels Principal Component Analysis (PCA) in terms of relative error ratio, the magnitude of objective function values, and the obtained CNOP pattern. Besides, compared to deepFE-based Particle Swarm Optimization, PEBOA has better solving efficiency with 3.4% larger objective function values and 2.2 times greater likelihood of obtaining the valid CNOP. Furthermore, the modified acquisition function reduces computation time by about 71.0% with four cores. The physical mechanism analysis shows that CNOPs obtained by PEBOA are almost identical to the adjoint method, which can cause an anomalous increase (decrease) in upstream Kuroshio transport and maintain physics consistency.
关键词	Autoencoder based Bayesian optimization Additive Gaussian kernels Paralleled acquisition function CNOP Regional ocean modeling system Upstream Kuroshio transport
DOI	10.1016/j.ocemod.2023.102213
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation of China[42075141]; National Natural Science Foundation of China[U2142211]; Key Project Fund of Shanghai 2020 Science and Technology Innovation Action Planfor Social Development[20dz1200702]; National Key Research and Development Program of China[2020YFA0608000]; first batch of Model Interdisciplinary Joint Research Projects of Tongji University[YB-21-202110]
WOS研究方向	Meteorology & Atmospheric Sciences ; Oceanography
WOS类目	Meteorology & Atmospheric Sciences ; Oceanography
WOS记录号	WOS:001010831800001
出版者	ELSEVIER SCI LTD
WOS关键词	NONLINEAR OPTIMAL PERTURBATION ; PARTICLE SWARM OPTIMIZATION ; PRINCIPAL COMPONENT ANALYSIS ; PACIFIC SUBTROPICAL COUNTERCURRENT ; NORTH EQUATORIAL CURRENT ; INTERANNUAL VARIABILITY ; TARGETED OBSERVATIONS ; SENSITIVE AREAS ; KUROSHIO ; PREDICTION
引用统计
文献类型	期刊论文
条目标识符	http://ir.qdio.ac.cn/handle/337002/182226
专题	海洋环流与波动重点实验室
通讯作者	Mu, Bin
作者单位	1.Tongji Univ, Sch Software Engn, Shanghai 201804, Peoples R China 2.Chinese Acad Sci, Inst Oceanol, CAS Key Lab Ocean Circulat & Waves, Qingdao 266071, Peoples R China 3.Laoshan Lab, Qingdao 266071, Peoples R China
推荐引用方式 GB/T 7714	Yuan, Shijin,Liu, Yaxuan,Qin, Bo,et al. A paralleled embedding high-dimensional Bayesian optimization with additive Gaussian kernels for solving CNOP[J]. OCEAN MODELLING,2023,184:26.
APA	Yuan, Shijin,Liu, Yaxuan,Qin, Bo,Mu, Bin,&Zhang, Kun.(2023).A paralleled embedding high-dimensional Bayesian optimization with additive Gaussian kernels for solving CNOP.OCEAN MODELLING,184,26.
MLA	Yuan, Shijin,et al."A paralleled embedding high-dimensional Bayesian optimization with additive Gaussian kernels for solving CNOP".OCEAN MODELLING 184(2023):26.