A Machine Learning Approach to GDP Prediction by Analyzing Economic Indicators

A country's economy is important for its growth and development. A well-performed economy describes the quality living of the citizens. For assessing the health of nation's economy the GDP performs a crucial benchmark. The country's finished goods total value and services in a particular period are called GDP. The GDP value of the country tells whether that country is developed country or developing country. If the GDP grows the common man of the country earns the quality lifestyle. It makes the country to generate more tax revenue. This paper deals with the GDP prediction of the country by creating the machine learning model in python jupyter notebook. The machine learning model predicts the GDP based on the Life Expectancy, Human Development Index (HDI), CO2 emissions per person and percentage of service workers in the country. It uses the gapminder dataset from the Kaggle repository. Diverse machine learning algorithms are employed in the process of construction of a machine learning model. We picked the algorithm by analyzing the structure of dataset that have taken for the model building. Given that the dataset is in the form of numerical values and arranged in a non-linear structure, the non-linear regression algorithms were to implement this system. The Average Absolute Error (AAE), Squared Error (SE) and R-squared (R 2 ) metrics were taken to validate the accuracy of the system. Among the three models that were implemented and tested Decision Tree Regression model (DTR), Support Vector Regression model (SVR) and Polynomial Regression model (PR), the Decision Tree Regression model achieved the error Free State and R-squared value of 1 which is the most desirable model.