Performance Measures for Evaluating the Accuracy of Time Series Hybrid Model Using High Frequency Data

Given that the traditional ARIMAX model has rarely been applied to any of climate change and environmental agents, which are most cognate agents with associated exogenous variables; neutralize for a better enhanced prediction system, distributional form error term is robust sufficient in capturing accommodating both external covariate(s) high frequency data required. This study therefore evaluates forecasting accuracy two models namely log-ARIMAX. The monthly adjusted recorded by four Oil Gas companies from 2005 – 2020 were used. forecastability was evaluated different matrices. effect Akaike Information Criterion (AIC) linear correlation on candidate among considered oil spill tested discussed. Results LOG-ARIMAX Models selection respect AIC show log-ARIMAX more efficient performed than observations characterized kurtosis, skewness, outliers, large fluctuation series heavy tailed traits as seen data.