Penerapan Stochastic Gradient Descent Support Vector Regression pada Data Laju Pertumbuhan Produk Domestik Bruto di Indonesia

Abstract

Economic growth is defined as an increase in a country's income as measured using gross domestic product (GDP) data. The GDP growth rate tends to experience an upward trend even though fluctuates in each time period. This fluctuation will affect the decision of investors in investing or withdrawing capital. In this study, a regression model is applied to GDP growth rate data to help investors understand the pattern of GDP growth in the future. One of the regression models that can be used is support vector regression with stochastic gradient descent optimization algorithm (SGD-SVR). The results show that the SGD-SVR model is able to be applied to GDP growth rate data in Indonesia. In the training process, the MSE resulted was 0.2805 with a total of 360 iterations. Meanwhile, the testing process resulted in an MSE of 0.0325.