Paper Title
Forecasting of Solar Radiation by Using Singular Spectrum Analysis Method

Abstract
The objective of this study is to predict the solar radiance in Langkawi by using Recurrent Forecasting-Singular Spectrum Analysis (RF-SSA) and Vector Forecasting-Singular Spectrum Analysis (VF-SSA). The study also investigates to present the comparison between two forecasting algorithms. The data used in this study is solar data for year 2020. As part of its methodology, this study had used an imputation method for handling missing data. Mean imputation, median imputation and mode imputation are three model used to obtain the missing value. The proposed model was analysed by using the performance measures that are Root Mean Square Error (RMSE). The result of imputation methods shows mode imputation is having the lowest RMSE value. The complete data sets of solar radiation then conducted by using Singular Spectrum Analysis(SSA). In the conclusion of this study, the comparison of results from the RF-SSA and VF-SSA reveals that recurrent forecasting reproduces the most reasonable hourly solar radiation and making it a perfect candidate for solar prediction research. The implication of this study is to establish the best model in Singular Spectrum Analysis(SSA) in order to obtain accuracy predicted value for solar radiance. Keywords - solar radiance; Singular Spectrum Analysis (SSA); Recurrent Forecasting (RF); Vector Forecasting (VF); forecasting; imputation