Share this post on:

Alue or monitored value, Xr is definitely the predicted worth or monitored worth before the normalization, Xmax could be the maximum predicted value or monitored value ahead of the normalization, and Xmin could be the minimum predicted value or monitored worth ahead of the normalization.Appl. Sci. 2021, 11,12 ofFigure 5. Comparison in between the HC-LSSVM model and also other investigation final results.four. Conclusions (1) On the basis in the leave-one-out cross-validation strategy, the Homotopy continuation method was utilised to optimize the LSSVM model parameters with the aim of minimizing the sum of squares of your prediction errors from the complete sample retention one particular, then the HC-LSSVM model was constructed, which solved the problems of low search efficiency in the search method and lack of global optimal remedy within the search outcomes on the current LSSVM models. Comparing with training samples and test samples, the HC-LSSVM model can accurately predict soft soil settlement, as well as the prediction result is substantially better than that of ordinary LSSVM model. The analysis outcomes give a new approach for the prediction of soft soil settlement. The prediction of future settlement quantity determined by the current observation data can successfully prevent the GS-626510 manufacturer occurrence of disasters.(2)(3)Author Contributions: Conceptualization, C.Z. and Z.L.; methodology, Z.L.; application, G.C. and S.X.; validation, Z.L. and G.C.; formal evaluation, Z.L. and G.C.; investigation, G.C. and S.X.; sources, C.Z. and Z.L.; information curation, G.C. and S.X.; writing–original draft preparation, G.C. and S.X.; writing– review and editing, G.C. and Z.L.; visualization, G.C. and Z.L.; supervision, G.C. and Z.L.; project administration, C.Z.; funding acquisition, C.Z. All authors have study and agreed to the published version from the manuscript. Funding: This research was funded by the National Key Analysis and Development Project, Grant Quantity 2017YFC1501203 and 2017YFC1501201; the National All-natural Science Foundation of China (NSFC), Grant Number 41977230; and the Unique Fund Important Project of Applied Science and Technology Study and Improvement in Guangdong, Grant Number 2015B090925016 and 2016B010124007. Institutional Evaluation Board Statement: Not applicable. Informed Consent Statement: Not applicable. Information Availability Statement: The information presented in this study are readily available inside the post. Acknowledgments: The authors would prefer to thank the anonymous reviewers for their extremely constructive and valuable comments.Appl. Sci. 2021, 11,13 ofConflicts of Interest: The authors declare no conflict of interest.Abbreviationsxi Rm yi Rn n H w b C ek i S(p) S(p- ) A-1 (p,p) A-1 (p- ,p) two K(xk , xl ) K (p,p- ) t_step C_step sig2_step f C sig2 Xn Xr Xmax Xmin L e n Cv wn wL k OCR Cc Cr E qu h Cst Csa Av Pv a Npv Nav Input vector Input space ML-SA1 Protocol Output vector Output space Number of coaching samples Kernel space mapping function Feature space Weight vector in space H Offset parameter Tunable regularization parameter Error variables Lagrange multiplier The p th element in S Column vector of S minus the p th element Element in row p and column p of A-1 Column vector of the column p of A-1 minus the p th element Kernel function parameter, labeled sig2 Dot item kernel function Row vector in the row p of K minus the p th element Homotopy parameter step size Regularized parameter step size Kernel function parameter step size Mapping of input space to output space Tunable regularization parameter of homotopy continuation system Kernel functi.

Share this post on:

Author: haoyuan2014