ResearchBib Share Your Research, Maximize Your Social Impacts
Sign for Notice Everyday Sign up >> Login

INVESTIGATING THE POTENTIAL INCREASE IN SEA LEVEL AT THE EAST COAST OF PENINSULAR MALAYSIA: TERENGGANU BY UTILISING VARIOUS MACHINE LEARNING TECHNIQUES

Journal: International Journal of Civil Engineering and Technology (IJCIET) (Vol.10, No. 6)

Publication Date:

Authors : ;

Page : 181-193

Keywords : Genetic Programming (GP); Predicting model; Sea Level (SL); Support Vector machine (SVM);

Source : Downloadexternal Find it from : Google Scholarexternal

Abstract

The inundation impact of sea level rise (SLR) is critical also severely affect the coastal regions of many countries and the health and safety of communities. In order to figure out the adaptation measures, this has drawn much research interest and attention and various artificial intelligence techniques have been simulated in previous studies to examine the performance of climate modelling. One of the significant drawbacks in current SLR impact studies is due to the lack of reliable methods for simulating sea level rise for the area of Kerteh, Terengganu, Malaysia. Therefore, this article aims to investigate the potential increase in sea level at the study area of Kerteh, Terengganu, Malaysia by comprising the model performance R and RMSE in order to measure the reliability of the prediction on the SVM and GP models. Two evaluation processes were used to determine the performance of the model. The initial assessment was based on partitioning the data into two sets namely; Monthly (Scenarios 1) and Seasonally (Scenario 2). The second and third assessment processes determined the most effective input to construct the models using a single and various possible combination of dependent and independent parameters respectively with the simulation adopted is SVM Regression (RSVM) with Pearson Universal Kernel (PUK). In terms of R was 0.837 with Scenario 2 outperformed Scenario 1. Overall, SVM is outperformed GP with the consistency of generalisation and time execution.

Last modified: 2019-08-09 14:16:42