Gaussian Process Regression Method for Exploring Economic Growth for Nigeria under the violation of Linearity Assumption

In this study, the estimates and the contribution of the identified economic drivers was oneconomic growth rate (RGDP) was considered when linearity assumption was violated. A gaussianprocess regression method was employed to quarterly data (1986 -2021) extracted from Nigeria’s ApexBank. Consequently, to obtain optimal and stable parameter estimate for the efficient prediction ofeconomic growth (RGDP), a gaussian process regression technique was adopted. The magnitude of thegaussian process estimated parameter for INDT, EXDT, RINR, REXR and OPEN were 0.4969, 0.1140,0.7424, 3.7881 and 3.1143 respectively. The sensitivity which is the expected contributions of theaforementioned economic growth drivers were revealed to 59.03%, 10.12%, 7.56%, 36.71% and 2.45%respectively in determining economic growth. However, from the gaussian process model, the main effector contributions of the economic growth drivers to the RGDP were 38.53%, 0.21%, 1.37% and 0.49% forINDT, EXDT, RINR, REXR and OPEN respectively. In this study, the result indicated that 6.63% of INDT,0.17% of OPEN, 6.66% of EXDT, 4.3175 of REXR and 3.1051% of RINR predicted 10.33% growth inRGDP at the desirability value of 0.46. Therefore, it can be concluded based on the findings thatgaussian process regression method was appropriate model for estimating and predicting the state ofeconomy under the violation linearity assumption. Also, government and policy makers must properlyharness the benefit of trade openness to grow the economy. The development of infrastructure and thegrowth of the economy through borrowing either internal or external are not sustainable enough, due tothe exchange rate volatility and high interest rate and the need for policy direction to address the negativeimpact of the identified monetary policy instrument toward economic growth enhancement.