A Regression Model-Based Approach to Identifying Determining Factors for GDP Growth in Bangladesh

GDP per capita is a key indicator for every country to measure the economic growth and standard of living. Recently, Bangladesh has gained a prospective rise in GDP per capita and it is, therefore, worth identifying influential factors associated with such an increase. This paper investigates various factors to find their significance towards this GDP growth and builds a prediction model to forecast the future. The significance analyses and model building processes are conducted on the World Bank data repository, based on the World Development Indicator (WDI)-2019. In this paper, we use multiple stepwise linear regression processes to build various models at first attempt, and their performances are measured using Mean Absolute Error (MAE), Root Mean Squared Error (RMSE) and Relative Absolute Error (RAE) under the shed of k-fold cross-validation technique. Later, we compare these results with other popular classifying techniques such as Gaussian Process, Decision Table, Random Tree, Multilayer Perceptron and Random Forest. Finally, the regression-based model yields significant performance with greater accuracy in forecasting the GDP per capita of Bangladesh and provides insight into the determinants for the potential increase.