Algorithm and hyperparameter optimizations for hetero-device classification by near-infrared spectra of falsified and substandard amoxicillin capsules.

In this work, we optimized classification algorithms and the hyperparameters for screening falsified and substandard amoxicillin capsules. The distribution of low-quality medical products is a serious problem, especially in low- and middle-income countries. Near-infrared (NIR) spectroscopy has been proposed as the first choice for a screening device. However, preparation of the reference library for the classification training is a highly difficult process. We herein propose a hetero-device classification between training and test devices. In this proposal, Fourier-transform NIR spectrometer and portable wavelength dispersive NIR spectrometer were used as training and test devices, respectively. As the classifier candidates, we examined 13 algorithms and selected 8. We then optimized the hyperparameters for these classifiers by the grid search and cross validation methods. In the final analysis, few classifiers were found to give acceptable prediction results by the hetero-device classification. When using these methods, it is crucial to examine the results by the classification probability, due to the trade-off between sensitivity and specificity. Finally, we suggest that k-nearest neighbors, extra trees, and gradient boosting classifiers are the optimal algorithms with high classification probability for the substandard and falsified amoxicillin capsules.

Device-Independent Discrimination of Falsified Amoxicillin Capsules Using Heterogeneous Near-Infrared Spectroscopic Devices for Training and Testing of a Support Vector Machine.

The objective of this work is to demonstrate the potential of near-infrared spectroscopy for common screening of falsified medicines in the field by means of a device-independent universal discrimination approach. In order to provide a useful discrimination tool to protect people from low-quality medical products, not only is a low-cost and portable screening device necessary, but a reference library is also essential. The authors believe that a device-dependent reference library inhibits near-infrared spectroscopy from becoming a popular screening tool. In this study, to develop a device-independent method, discrimination performance is evaluated using different devices for training and testing. The training data sets for the reference library were prepared using a bench-top Fourier transform near-infrared spectrophotometer, and predictive discrimination was performed using the spectral data by a low-cost and portable wavelength dispersive near-infrared spectrophotometer. A near-infrared spectrum-based support vector machine was used for these purposes, but the screening resulted in low accuracy thought to be caused by the intrinsically device-dependent features of the spectra data. Thus, principal component analysis was performed to collect the proper components to discriminate low-quality products from standard products. The principal component score-based support vector machine was able to produce highly accurate results, identifying falsified products with no false positive cases.