Skill Comparison of Multiple-Linear Regression Model and Artificial Neural Network Model in Seasonal Rainfall Prediction-North East Nigeria
Ebiendele Ebosele Peter *
Department of Basic Science, Auchi Polytechnic, Auchi, Nigeria.
Ebiendele Eromosele Precious
Department of Meteorology and Climate Science, Federal University of Technology, Akure, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
Seasonal Rainfall in Nigeria is usually complex due to the tremendous range of variation over a wide range of scales both in space and time. Forecasting techniques such as Multiple Linear Regression (MLR) and Artificial Neural Network (ANN) have been used to study rainfall. This study is motivated by the need to compare MLR and ANN models to know which one is more reliable in predicting Seasonal rainfall. The rainfall datasets used in this study were collected from the achieving of Nigeria meteorological agency from (1986-2017). The model comparison was based on four criteria; the Mean Absolute Error (MAE), Root Mean Square Error (RMSE), prediction error and Correlation coefficient (r). The error measures are comparable for the two models. The analysis of the models performance shows that, overall, the ANN model performs better than the MLR model in terms of PE, RMSE, MAE and correlation coefficient. ANN had the minimum MAE=56.66 mm, RMSE=74.84 and PE=0.11209 respectively and in terms of correlation coefficient between the observed rainfall and predicted rainfall amount, ANN model had high correlation coefficient (0.93) compared to MLR model whose correlation coefficient was (0.66).
Keywords: Prediction, multiple regression, artificial neural network, seasonal rainfall