Downloaded by 74.47.189.78 on October 29, 2015 | http://pubs.acs.org Publication Date (Web): October 7, 2015 | doi: 10.1021/bk-2015-1199.pr001
Preface This American Chemical Society (ACS) Symposium series text is based on the full day symposium entitled, “The Birth of Chemometrics - In Honor and Memory of Bruce Kowalski,” that was held at FACSS in Milwaukee, WI (October 2013) and cosponsored by the Division of Computers and Chemistry (COMP) of the American Chemical Society. Bruce Kowalski is recognized by the scientific community as the founder of the field of chemometrics. This symposium Series text is a follow up to the Symposium Series Volume 52 (Chemometrics: Theory and Application), edited by Bruce Kowalski and is based on the symposium organized by Bruce at the National ACS meeting in San Francisco in 1976, which was also cosponsored by COMP. The 14 contributors to the current volume (see Table of Contents) are all leaders in the field of chemometrics and have strong personal recollections of Bruce as a man who was a catalyst able to bring together creative minds. All major areas in the field are well represented in this collection: pattern recognition, library searching, multivariate calibration, multivariate curve resolution, variable selection, data fusion, calibration transfer, environmental chemometrics, forensics, and biological and mixture analysis. Many chapters have a link to previous work done by Bruce and will serve as a retrospective to the career of Bruce Kowalski, who believed that a rational approach was needed to improve both the quality of measurements and to extract information from them. Bruce believed that chemometrics would serve as a guiding theory for analytical chemistry and believed that it would be used both to optimize existing analytical methodology and to direct researchers attempting to construct better tools. Each chapter in this text demonstrates the progress that has been made in the field towards the realization of Bruce Kowalski’s goal. The first chapter of the text entitled, “Chemometrics and Bruce: Some Fond Memories,” is written by Svante Wold and describes the history of empirical and semi-empirical “data driven, soft, analogy” models for the design of experiments and the analysis of the resulting data. This history is marked by a number of influential events inspired and encouraged by Bruce and is illustrated by examples of method development driven by necessity to solve specific problems and leading to data driven soft models, which have been shown to be superior to the classical first principles approaches to the same problems. Although Bruce is recognized for his accomplishments in attracting the talents of chemists and engineers, he also showed enormous vision in his efforts to assimilate statisticians and mathematicians into the field of chemometrics. Bill Rayens, a mathematician turned statistician, describes his journey from statistician to chemometrician as a result of his interactions with Bruce Kowalski ix In 40 Years of Chemometrics – From Bruce Kowalski to the Future; Lavine, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2015.
Downloaded by 74.47.189.78 on October 29, 2015 | http://pubs.acs.org Publication Date (Web): October 7, 2015 | doi: 10.1021/bk-2015-1199.pr001
in Chapter Two. Chapter Three, written by Peter Wentzell, describes the evolution of maximum likelihood principal component analysis and related techniques from a personal perspective highlighting the author’s collaboration with Bruce Kowalski. Chapter Four focuses on Bruce Kowalski as a mentor, innovator and pioneer through the solution to a problem involving dioxin involving dioxin concentrations in excess of background levels in the harbor of Port Angeles in Western Washington. A mixture analysis study undertaken by Scott Ramos using pattern recognition and multivariate curve resolution methods to understand the nature of the contamination indicated several characteristic patterns that could be associated with identifiable source materials. The work in this study was performed at Infometrix, a software company founded by Bruce Kowalski in 1978. Chapters Five and Six focus on multivariate curve resolution. In Chapter Five, Roma Tauler provides an exhaustive review of multivariate curve resolution, which is the generic denomination for a family of methods used to solve the ubiquitous problem of mixture analysis. Tauler became interested in curve resolution while visiting Bruce in the late 1980’s. Phil Hopke in Chapter Six describes some recent developments in multivariate curve resolution related to the problem of source apportionment in air monitoring of atmospheric particulates. Recent developments in the field of pattern recognition are delineated by Steve Brown and Barry Lavine in Chapters Seven and Eight. In Chapter Seven, hierarchical class modeling approach in which samples receive more than one class label are compared to traditional “flat” classification for the modeling of hierarchical geospatial data, a problem that relates to those studied by Brown and Kowalski in the late 1970’s. In Chapter Eight, pattern recognition techniques are applied to the problem of searching the infrared spectral libraries of the Paint Data Query (PDQ) automotive paint database to differentiate between similar IR spectra and to determine the assembly plant, model, and line of an automotive vehicle from a clear coat paint smear recovered at a crime scene where damage to a vehicle and/or injury or death to a pedestrian has occurred. This, too, echoes work started by Brue in early studies of pattern recognition, but his focus was on paper. Chapters Nine, Ten, and Eleven focus on recent developments in multivariate calibration, another area where Bruce contributed significantly. Model selection is usually limited to the evaluation of cross validation prediction errors. However, there are advantages of using multiple criteria for model selection, which is discussed by John Kalivas in Chapter Nine. Chapter Ten focuses on the solution to the variable selection problem in PLS and PCR using adaptive regression subspace elimination approach pioneered by Karl Booksh. The essentials of multivariate calibration transfer are discussed by Jerry Workman in Chapter Eleven. The remaining three chapters of this text focus on biological applications of chemometrics. The field of proteomics and metabolomics from the standpoint of chemometrics is reviewed by Jeff Cramer in Chapter Twelve. Rene Jiji in Chapter Thirteen explores the application of data fusion for spectroscopic data to improve predictions of protein secondary structure. The remaining chapter x In 40 Years of Chemometrics – From Bruce Kowalski to the Future; Lavine, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2015.
Downloaded by 74.47.189.78 on October 29, 2015 | http://pubs.acs.org Publication Date (Web): October 7, 2015 | doi: 10.1021/bk-2015-1199.pr001
by Frank Vogt summarizes nonlinear modeling of microalgal biomasses for the purpose of exploring the impact of pollutants on our environment. This text will be of interest to individuals who are interested in modeling data. Interest in modeling data continues to grow with the emergence of new areas such as computational statistics, business intelligence, big data, and analytics. In chemistry, modeling of data has taken a different path as it has become integrated into the field of analytical chemistry. Because chemometrics is not well understood by chemists, this text should prove beneficial and be of great interest to researchers who need to take advantage of techniques such as principal component analysis, partial least squares, linear discriminant analysis and outlier analysis in their work. This book allows the reader quick access to different areas of current research in chemometrics featured in the literature by providing key references and viewpoints not found elsewhere. This text also highlights changes that have occurred in the field since its origins in the mid-1970’s and will serve as a report on the current state of the art of the field of chemometrics. The editors of this text believe that it will be of interest not only to physical scientists and engineers but also to statisticians and informatics types who have come to the realization that chemometrics is worth a second look.
Barry K. Lavine Oklahoma State University Department of Chemistry Stillwater, Oklahoma 74078
Steven D. Brown University of Delaware Department of Chemistry and Biochemistry Newark, Delaware 19716
Karl S. Booksh University of Delaware Department of Chemistry and Biochemistry Newark, Delaware 19716
xi In 40 Years of Chemometrics – From Bruce Kowalski to the Future; Lavine, et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2015.
