Analisis Kandungan Minyak Babi Pada Minyak Kanola Melalui Klasifikasi Pola Hidung Elektronik (E-Nose) Berbasis Linear Diskriminan Analysis(LDA)

Abstract

Listen

Translate article

The classification of the aroma pattern of pork oil and canola oil on the chemometric-based electronic nose has been classified. The electronic nose used is a series of chemical sensors combined in parallel. Sensors are made of semiconductor material that can detect changes in gas in the air. Each sample measured by an electronic nose provides output in the form of different voltages on each sensor. The data processing method used is Linear Discriminant Analysis (LDA) which is able to classify based on patterns. The samples used were canola oil, pork oil and a mixture of pork oil and canola oil with a percentage of 50%: 50%. The results of the classification of electronic nose patterns with samples of pork oil and canola oil show that each sample is fairly well clustered with the value of the first disk function is 99.9% and the second discriminant function value is 0.1%.