การเจริญเติบโตทางเศรษฐกิจจากการศึกษาในประเทศไทยโดยศึกษาเปรียบเทียบ วิธีกำลังสองน้อยที่สุดกับวิธิโครงข่ายประสาทเทียม / ยืนยง นิลสยาม = The Contribution of Education to Economic Growth in Thailand : least square approach and neural network approach / Yuenyong Nilsiam
The aim of this research is to compare the forecasting accuracy of least square and neural network methods for the contribution of education to economic growth in Thailand. Economic growth arises from many factors. The important factor is human, who is labor and entrepreneur. However, economic growth cannot be explained completely by human and capital. Many economists agree that quality of human or human capital is a significant factor for economic growth. Even though, human capital consists of education, health, and so on, economists have been interested in education. The obvious relationship of education and economic growth is a greater earning of labor with has higher education. This research use gross domestic product data, capital stock of Thailand data and labor with education level during 2518-2549. These data are used to build the models, least square method using Cobb-Douglas function, least square method using CES function, linear neural network, and nonlinear neural network. For each model, average errors are calculated. The results of the study show that the least square method using Cobb-Douglas function gives the most accurate average forecasting values for economic growth. Less accurately average forecasting values are given by the least square method using CES function and linear neural network, respectively. Lastly, nonlinear neural network gives the least accurate average forecasting value for economic growth.