Determination of Adulteration in Philippine Coconut Wine using Principal Component Analysis and Soft Independent Modelling of Class Analogy

Local alcoholic beverages such as “Lambanog” or Coconut wine started in 15th century and has been passed down through generations by coconut farmers garnering acceptance from global wine connoisseurs because of its sweet pungent smell, clean taste, smooth and long finish. Commercialization of local coconut wine increases the production of adulterated liquors that sets risks for the local distillery industry as well as the consumers' health. Examples of health risks include blindness, nausea and in some cases death. In this paper, the percentage of methanol as well as other compounds: Ethanol and Ethylene Glycol was able to obtain from a Near Infrared (NIR) spectrometer using a camera that captures the spectrum that passed through a sample, sends the data to a Raspberry Pi 3b which contains a software based on Principal Component Analysis (PCA) to classify the samples by their unique wavelength range and Soft Independent Modelling of Class Analogy (SIMCA) to further classify them into adulteration percentages. There are 40 samples of Coconut wine with varying adulteration ranging from 5 to 50 percent. With 95% confidence the population mean is between 77.2 and 85.9, based on 160 samples. Therefore the results of the proposed method provided statistically significant results in comparison to the reference method's result.