Intelligent techniques for modelling the effects of methanol leaves extract of Combretum hypopilinum diels (Combretaceae) on hepatic enzymes as indicators of radiation-induced liver injury

The need for the determination of hepatic enzymes such as alkaline phosphatase, aspartate aminotransferase, and alanine aminotransferase is of paramount importance owing to their significant contribution to metabolic functions and diverse bodily functions. In the present work, two different techniques in artificial intelligence namely, artificial neural network and support vector machine were utilized in the simulation. The independent variables for this prediction were total protein, direct bilirubin, total bilirubin, albumin and glucose. To assess how well the simulated values matched the observed values of the target hepatic enzymes, two evaluation metrics were employed namely, determination coefficient and Correlation co-efficient. Additionally, the mean square error and root mean square error were utilized to quantify the errors generated by each model combination. Consequently, a novel hybrid technique was proposed. The findings showed that ANN-PSO displayed higher prediction skills and boosted the efficacy of the single models.