Combining experimental design and artificial neural networks for the determination of chlorinated compounds in fish using matrix solid-phase dispersion

The purpose of this contribution is to present a very effective strategy for the development of matrix solid phase dispersion (MSPD) extraction methodology for the determination of chlorinated compounds in fish using experimental design methods and artificial neural networks (ANNs). The MSPD extraction method of compounds is a preparation method that comprises sample homogenization, cellular disruption, fractionation and purification in a single process. Many parameters have to be taken care of when developing an MSPD extraction method because its performance is mainly affected by column packing and elution procedure. In this study, the best possible performance of MSPD has been achieved using experimental design and ANN modeling. The ANN used is a multilayer perceptron (MLP) trained with the standard error back propagation algorithm. Experimental results demonstrate that the proposed soft computing strategy is very effective, efficient and achieves very satisfactory results.

► Development of matrix solid phase dispersion (MSPD) extraction methodology. ► Determination of chlorinated compounds in fish. ► Uses experimental design methods and artificial neural networks (ANNs). ► Multilayer perceptron (MLP) trained with standard error back propagation algorithm. ► Experimental results demonstrate effectiveness and efficiency.