صادق کریمی

In factor analysis-based regression and pattern recognition models, the latent variables (LV's) are calculated from the response data measured at all employed instrument channels. Since some channels are irrelevant and their responses do not possess useful information, the extracted LV's possess mixed information from both useful and irrelevant channels. Recently we have proposed a new variable selection algorithm, based on clustering of variable concept to solve the aforementioned problem. The basic idea behind the clustering of variable is that, the instrument channels are clustered into different clusters via clustering algorithms. Then, the spectral data of each cluster are subjected to PLS regression. In the present study the application of this variable selection technique has been evaluated in electrochemical data. A simple, selective and sensitive sensor based on mesoporous silica nanoparticles modified carbon paste electrode (MSNs/CPE) is introduced for simultaneous electrochemical determination of tyrosine (Tyr) and tryptophan (Trp) . Compared with the unmodified electrode, the oxidation peak current significantly improved for both amino acids. The results showed that the method was successfully applied to the simultaneous determination of Trp and Tyr in some synthetic and real samples. Under optimized conditions, the oxidation peak current of Trp was linear over a concentration range of 0.05 to 600 ?M with a detection limit of 1.13 × 10-8 M. The oxidation peak current of Tyr was linear over a concentration range from 0.3 to 600 ?M with a detection limit of 4.97 ×10-8 M. The obtained statistical parameter indicates that variable clustering can split useful part from redundant ones, and then based on informative cluster; stable model can be reached.