Golden Section Search-Multi Variable Algorithm for Optimization Parameter of Triple Exponential Smoothing Algorithm to Predict Sufferers of Lungs Disease

Triple Exponential Smoothing is a prediction algorithm that considers time series data pattern and trends. The algorithm performance is getting better inline with the data volume. The specific feature of this algorithm is smoothing the previous period's data before performs a prediction task. However, this algorithm has weakness in setting the initial value of three parameters which are alfa, beta, and gamma, and it influences the prediction's accuracy. Therefore, we suggest Golden Section Search (GSS) optimization algorithm. The standard version of GSS optimizes a single variable. We modified GSS in order to optimize three parameters and called Golden Section Search-Multi Variable (GSS-MV). In order to optimize the parameters an extension of GSSMV with the triple exponential smoothing (called GSSMV-TES) will be proposed and tested in this paper. In order to evaluate the performance of the proposed algorithm a dataset of used the lungs' disease collected from a publicly available data source. In this research, two testing scenarios are designed. In the first scenario the data divided into training and testing in equal percentage (50% of each). In the second scenario, the data divided into 60% and 40% for training and testing respectively. According to the experiment results, GSSMV-TES shows better performance in the first scenario with the mean absolute percentage error (MAPE) at 3%. The mean percentage error of GSSMV-TES is observed 4% better than triple exponential smoothing (TES).