数学
主成分分析
样品(材料)
模式识别(心理学)
统计
人工智能
化学
色谱法
计算机科学
作者
Meinilwita Yulia,Diding Suhandy
出处
期刊:IOP conference series
[IOP Publishing]
日期:2022-06-01
卷期号:1038 (1): 012035-012035
被引量:1
标识
DOI:10.1088/1755-1315/1038/1/012035
摘要
Abstract This current research presents a simple analytical method for classifying organic and conventional coffee samples from different origins. UV pre-processed spectral data in the range of 250-400 nm was used to discriminate between organic Lampung robusta coffee from Lampung Barat (n=50) and two conventional Lampung robusta coffees from Lampung Barat (n=50) and Tanggamus (n=50). Ground roasted coffee samples with 50 mesh were used for samples. UV-vis spectrometer was utilized to acquire UV spectral data from an aqueous coffee sample. A chemometric method based on PCA and PCA-LDA algorithm was used to classify the samples. The PCA result shows all organic coffee samples were clustered on the negative of PC1 while all conventional coffee samples were on the positive of PC1. The conventional coffee samples from Lampung Barat and Tanggamus were grouped in different clusters according to their origin. The PCA-LDA resulted in a 100% accuracy in classification both for calibration and prediction. This method is a promising approach for organic Lampung robusta authentication with a relatively low-cost spectrometer and simple sample preparation.
科研通智能强力驱动
Strongly Powered by AbleSci AI