Self-organizing maps based on chaotic parameters to detect adulterations of extra virgin olive oil with inferior edible oils

Abstract

A nonlinear algorithm based on chaotic parameters (CPs) has been employed to determine the nature of different output signals obtained from UV–vis spectrophotometer (UV) measurements. These signals come from UV scans of adulterated samples of extra virgin olive oil (EVOO) with refined olive oil or refined olive pomace oil, or from pure samples of EVOO with white random or sinusoidal white random noises. The data collected from this equipment was used to calculate CP values. Then, a self-organizing map was used to detect different types of signals. Using this method, the signals can be identified and classified into five groups depending on their type, the percentage of noise added, and the concentration of adulterant agents, with a misclassification rate of less than 1.3%.