Machine Learning–Based Detection of Olive Oil Adulteration Using BME688 Gas Sensor Matrix
İsmail Hakkı Parlak, Mehmet Mıllı, Nursel Söylemez Milli
Abstract
Abstract Olive oil, an inseparable part of Mediterranean culture, is valued in a vast geography due to its healthiness and deliciousness. Higher quality olive oil is exposed to adulteration by mixing cheaper vegetable oils into it due to its high cost and decreased production volume due to the drought experienced in recent years. Mixing with lower-quality olive oil or mixing vegetable or seed oils such as sunflower, corn, cottonseed, peanut, soybean, rapeseed, or poppy seed oils is considered the most common type of adulteration in olive oil. In this study, sunflower oil was mixed into extra virgin olive oil at different concentrations, and the gases emitted by the mixtures were measured using high-sensitivity BME688 sensors produced by Bosh Sensortec. Various machine learning models were then trained using the data obtained from the sensors. It was revealed that both the classification and regression models detected extra virgin olive oil adulteration with near-perfect accuracy.