Prediction of Microporous Aluminophosphate AlPO4 -5 Based on Resampling Using Partial Least Squares and Logistic Discrimination

Abstract

In this paper, Partial Least Squares (PLS) regression and Logistic Discrimination (LD) are employed to predict the formation of microporous aluminophosphate AlPO4 -5 based on the database of AlPO synthesis, which aims to provide a useful guidance to the rational synthesis of microporous materials as well as other inorganic crystalline materials. To deal with the problem of class imbalance, four guided resampling methods considering not only the between-class imbalance but also the within-class imbalance are proposed. Experimental results indicate that the presented methods are competent for predicting the formation of microporous aluminophosphate AlPO4 -5. Specially, compared with some existing resampling methods, our proposed resampling methods exhibit much better predictive results.

Keywords: Logistic discrimination; Microporous materials; Partial least squares; Resampling.