SINE COSINE ALGORITHM BASED NEURAL NETWORK FOR RAINFALL DATA IMPUTATION

The Sine Cosine Algorithm (SCA) is a relatively recent metaheuristic algorithm, drawing inspiration from the characteristics of trigonometric sine and cosine functions. SCA has been widely used to address diverse optimization challenges in several domains. The advantages of SCA can be attributed to its simple implementation, reasonable execution time, and adaptability to hybridize with other optimization methods easily. This chapter presents the ability of the sine cosine algorithm-based neural network (SCANN) to predict and optimize missing rainfall at different percentages of missing rates. These findings revealed the superior performance of the SCANN imputation method compared to the feedforward neural network (FFNN) method, indicating its suitability for efficiently filling missing values in the rainfall database.