Cluster Analysis for IR and NIR Spectroscopy: Current Practices to Future Perspectives

• DOI: 10.32604/cmc.2021.018517
• Publication Date: 2021-07-22T04:43:38Z
• Authors: Simon Crase; Benjamin Hall; Suresh N. Thennadil
• Keywords: 0202 electrical engineering, electronic engineering, information engineering; 02 engineering and technology

Supervised machine learning techniques have become well established in the study of spectroscopy data. However, unsupervised technique cluster analysis hasn’t reached same level maturity chemometric analysis. This paper surveys recent studies which apply to NIR and IR In addition, we summarize current practices contrast these with literature from pattern recognition domain. includes data pre-processing, feature extraction, clustering distance metrics, algorithms validation techniques. Special consideration is given specific characteristics typically high dimensionality relatively low sample size. The findings highlighted a lack quantitative evaluation for With this mind, propose an model or workflow specifically suited along pragmatic application strategy.