Abstract:
The objective of this research is to compare the efficiency of forecasting AAPL stock prices for Apple Inc., a computer business company. Apple revolutionized desktop computing in 2513 with the Apple Two and later the Macintosh in 2523. Presently, Apple is renowned for its hardware such as iMac, iPad, and iPhone, serving as investment indicators. Market overview data, gathered from the Investing website from January 1, 2561, to December 1, 2566, totaling 72 months, was utilized. The data was divided into two sets: a 60-month training set and a 12-month testing set. Statistical methods were employed, selecting forecasting models suitable for each dataset. These included the Auto Regressive Integrated Moving Average (ARIMA) model, the Exponential Smoothing Method (ES), and a Decomposition method. Model selection criteria were based on Mean Absolute Percent Error (MAPE) to ensure forecasting efficiency. Upon evaluating the forecasting methods, it was found that the Decomposition method demonstrated the highest accuracy, with the lowest MAPE value of 8.09. In contrast, the ARIMA and Exponential Smoothing methods yielded MAPE values of 9.29 and 16.5, respectively. Therefore, the Decomposition method is considered the most effective for stock price forecasting