Modeling of the Retardance in Citric Acid Coated Ferrofluids Using Stepwise Regression

Citric acid (CA) coated Fe₃04 ferrofluids (FFs) have been conducted for biomedical application. The magneto-optical retardance of FFs was measured by a polarimeter [1]. Optimization and multiple regression of retardance in FFs were executed by Taguchi method and Excel, and the F value of regression model was large enough [1]. However, the modeling of retardance executed by Excel is not in a systematic way. In stepwise regression, a stepping procedure which iteratively altering the model established at the previous step by adding or removing variables in accordance with the stepping criteria is used in order to add significant variables or remove insignificant variables [2]. The results of stepwise regression by MATLAB was shown in Fig. 1, the model had highly predictable ability owing to F of 2.55897e+7 and had a high correlation coefficient of one. The average absolute error of predicted reatrdances was just 0.0044%.