Synergizing machine learning and chemometrics with voltammetric fingerprinting for the rapid detection of Piper retrofractum adulteration.

Long pepper is a prominent medicinal plant, extensively utilized as a bioactive constituent in traditional Asian medicine, including in Ayurvedic, Traditional Chinese, and Indonesian herbal medicines. However, the adulteration of long pepper in powdered form not only causes economic losses but also poses a health risk to consumers. This study developed and compared the integration of voltammetric fingerprinting with partial least squares regression (PLSR), Lasso Regression, support vector regression (SVR), random forest (RF), and k-nearest neighbors (kNN) for the quantitative analysis of long pepper, particularly Java long pepper (Piper retrofractum) adulteration. Cyclic voltammetry was performed at a MWCNT-COOH/Chitosan nanofiber-modified glassy carbon electrode to obtain representative voltammetric fingerprints. Principal component analysis (PCA) of CV data was employed for data exploration. Based on the PCA results, the selected optimum dataset comprises the current responses within the potential range of 0 to +1 V and back to 0 V, which contributed most significantly to discrimination between Java long pepper and adulterant. The optimized voltammetric data were subsequently processed using chemometric and machine learning approaches for quantitative analysis. Among all evaluated methods, PLSR consistently achieved the highest predictive performance across all potential regions. This result highlights the suitability of latent-variable chemometric approaches for high-dimensional voltammetric data, which are characterized by strong multicollinearity among neighbouring potentials. Specifically, the PLSR model constructed using the optimized CV dataset outperformed other methods (R 2 = 0.9926, RMSEC = 0.0266, and RMSECV = 0.0293). These findings demonstrate that combining CV with chemometrics and machine learning provides a rapid, effective method for portable in situ monitoring of spice adulteration.