A review of partial least squares modeling (PLSM) for water quality analysis

Abstract

Regression is a powerful tool in statistical modeling suited for qualitative and quantitative analysis and widely used in forecasting and prediction. The partial least squares modeling (PLSM) is one of the regression tools used in statistical analysis. There are many fields in which PLSM has been used; water is one of them, which is an area of interest for many researchers and scientists for more than two decades. Since water has multiple parameters to analyze, there is a problem of dimensionality and collinearity. The problem of multidimensionality, as well as collinearity, can be solved by PLSM. PLS regression can be suitable for analysis as it is the most prominent multivariate regression tool. This paper describes the use of PLS regression modeling for water quality analysis of different kinds of water samples (groundwater, wastewater, river water, and coastal water). Various methods employing PLSM for water quality analysis has been discussed in detail.