Publicação
Simultaneous determination of chlorpyrifos and carbaryl by spectrophotometry and boosting partial least squares
| Resumo: | Residue of pesticide has posed a serious threat to human health. Fast detection methods are necessary for the various types of pesticide. This study investigated the feasibility of the combination of spectrophotometry and boosting partial least squares method (boosting-PLS) for the simultaneous determination of chlorpyrifos and carbaryl, two of the most important pesticides in agriculture. Effect of pH on the absorbance spectra of each component was studied and pH 5 was selected as optimum. To build boosting-PLS models, thirty six binary mixtures were used and to validate the resulted models, twenty binary mixtures were employed. Results for boosting-PLS were compared to ones obtained for full-spectrum PLS and PLS algorithm coupled with tabu search (TS) as wavelength selection tool. In spite of the improvement by both boosting and wavelength selection approaches, the prediction ability of PLS algorithm, the boosting-PLS is superior to the other models. Boosting-PLS is not only feasible but also can avoid doing wavelength selection in such a task. |
|---|---|
| Autores principais: | Zare-Shahabadi,Vali |
| Outros Autores: | Abbasitabar,Fatemeh; Akhond,Morteza; Shamsipour,Mojtaba |
| Assunto: | chlorpyrifos carbaryl partial least squares tabu search boosting regression |
| Ano: | 2013 |
| País: | Brasil |
| Tipo de documento: | artigo |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | IBICT |
| Idioma: | inglês |
| Origem: | Oasisbr |
| Resumo: | Residue of pesticide has posed a serious threat to human health. Fast detection methods are necessary for the various types of pesticide. This study investigated the feasibility of the combination of spectrophotometry and boosting partial least squares method (boosting-PLS) for the simultaneous determination of chlorpyrifos and carbaryl, two of the most important pesticides in agriculture. Effect of pH on the absorbance spectra of each component was studied and pH 5 was selected as optimum. To build boosting-PLS models, thirty six binary mixtures were used and to validate the resulted models, twenty binary mixtures were employed. Results for boosting-PLS were compared to ones obtained for full-spectrum PLS and PLS algorithm coupled with tabu search (TS) as wavelength selection tool. In spite of the improvement by both boosting and wavelength selection approaches, the prediction ability of PLS algorithm, the boosting-PLS is superior to the other models. Boosting-PLS is not only feasible but also can avoid doing wavelength selection in such a task. |
|---|
Atividades financiadas
Não foram encontrados projetos financiados.