An EMD–PSO–LSSVM Hybrid Model for Significant Wave Height Prediction

The accurate prediction of significant wave height (SWH) offers major safety improvements for coastal and ocean engineering applications. However, the phenomenon is nonlinear nonstationary, which makes any work a non-straightforward task. aim research presented in this paper to improve predicted via hybrid algorithm. Firstly, an empirical mode decomposition (EMD) used preprocess data, are decomposed into several elementary signals. Then, least squares support vector machine (LSSVM) with learning ability adopted predict SWH, particle swarm optimization (PSO) automatically performs parameter selection LSSVM modeling. results show that EMD–PSO–LSSVM model can compensate lag timing models. Furthermore, performance has been greatly improved deep-sea area; accuracy coefficient determination (R2) increases from 0.991, 0.982, 0.959 0.993, 0.987, 0.965, respectively. proposed better than EMD–LSSVM Therefore, provides valuable solution SWH.