HALK: A hybrid active-learning Kriging approach and its applications for structural reliability analysis

AOK‐ES: Adaptive optimized Kriging combining efficient sampling for structural reliability analysis

“…[30][31][32] The core thought of adaptive Kriging (AK) is to adaptively improve the fitting precision of the LSF, by designing a learning function to acquire high-quality training samples. 33 Given the significant influence of learning function, a large number of Kriging-based learning functions have been presented, such as the least improvement function, 34 reliability-based expected improvement function, 35 weight learning function, 36 improved U function, 37 and hybrid learning function, 38 H-learning function. 39 Furthermore, efficient variance-reduced sampling techniques have been embedded into the adaptive reliability analysis field to improve computing efficacy.…”

A novel two-stage reliability analysis method combining improved cross-entropy adaptive sampling and relevant vector machine

“…In recent years, metamodel-based reliability analysis methods have received significant attention. Common metamodels used in reliability analysis include neural network [16], support vector machine (SVM) [17,18], response surface method [19,20] and Kriging [21][22][23]. From a machine learning perspective, the more samples used for fitting the surrogate model, the higher the accuracy of the surrogate model becomes.…”

On the impact of prior distributions on efficiency of sparse Gaussian process regression

“…Gaussian Processes (GP) is a generic supervised learning method designed to solve a wide range of probabilistic Machine Learning, including classification, regression, probabilistic forecasting, uncertainty quantification, etc [1,2,3,4]. The Gaussian processes regression (GPR) has been proven to be a powerful and effective method for non-linear regression problems due to many desirable properties such as simple to implement, flexibility and fully probabilistic models [5,6,7,8].…”