Pattern recognition in chemometrics

•Early identification of chemometrics with chemical pattern recognition•Pioneering work in the 1960s and 1970s•Change in emphasis of chemometrics in the 1980s•Loss of input from the statistics and machine learning community•Misunderstanding of commonly employed pattern recognition techniques

The origins of chemometrics within chemical pattern recognition of the 1960s and 1970s are described. Trends subsequent to that era have reduced the input of pattern recognition within mainstream chemometrics, with a few approaches such as PLS-DA and SIMCA becoming dominant. Meanwhile vibrant and ever expanding literature has developed within machine learning and applied statistics which has hardly touched the chemometric community. Within the wider scientific community, chemometric originated pattern recognition techniques such as PLS-DA have been widely adopted largely due to the existence of widespread packages, but are widely misunderstood and sometimes misapplied.