“…Least squares regression analysis is typically used for mapping samples of known compositions (Fu, Holtom, Freudiger, Zhang, & Xie, 2013; Lee, Moon, Migler, & Cicerone, 2011). For samples with little or no prior knowledge, methods such as principal component analysis (Lin et al, 2011), k‐means clustering (Krafft et al, 2009), spectral phasor analysis (Wei et al, 2019), Independent components analysis (Ozeki et al, 2012), multivariate curve resolution analysis (Zhang et al, 2013), as well as in‐house‐developed algorithm (Masia, Karuna, Borri, & Langbein, 2015; Masia, Pope, Watson, Langbein, & Borri, 2018) have been employed for mapping major components in Raman imaging data. Lately, machine learning (ML) algorithms have been adopted and developed to extract information from Raman spectra (mainly based on pre‐known knowledge), such as algorithms developed for assessing expressed human meibum (Alfonso‐Garcia et al, 2017), identification of pathogenic bacteria (Ho et al, 2019), and detecting prostate cancer (Lee, Lenferink, Otto, & Offerhaus, 2020).…”