Intelligent ocean wave height prediction system using light GBM model

Abstract

Forecasting the heights of marine waves is an important tool for offshore and coastal engineering and a huge undertaking in marine detection and warning. However, a precise forecast of the Sea Wave Height (SWH) is challenging and outstanding to waves' volatility and fluctuation characteristics. Therefore, our research proposes an Intelligent Ocean Wave Height Prediction system using a light gradient boosting machine learning. Wave speed, peak wave direction, zero up crossing wave period, wave period, and SWH are among the wave-based properties we extract. Then the inputs are fed into the Light GBM, which performs these high-dimensional inputs admirably, and the model is simple to interpret. Furthermore, because LightGBM is noise-insensitive and can work with unnecessary data in time-window-size data, the proposed method can be used to estimate wave height. Consequently, our proposed approach outperforms when compared to the existing techniques.