Classification and Comparison of Methods

In previous chapters of this book, we established a number of close basic relationships between groups of separation methods. We proceeded on the premise that the recognition and study of common features of scattered techniques are most worthwhile it facilitates understanding, aids evaluation, allows comparison, provides a basis for prediction, and simplifies the theory to a set of common elements. However, the further codification of like and unlike properties requires a system of classification. Unfortunately, the development of a complete classification system for separations is made difficult by the great number of variables of interest and their imperfect correlation with one another. The problem has been discussed in a paper by the author here we borrow liberally from that work [1]. 

This chapter begins with three sections on basic concepts of classification. In the last four sections of the chapter, we describe the role of flow in separations. This subject relates also to classification because flow is one of the major variables distinguishing one class of separation methods from another. 

---

Bowman, D., Gabbard, J., Hix, D. A Survey of Usability Evaluation in Virtual Environments Classification and Comparison of Methods. Presence Teleoperators and Virtual Environments 11(4), 404-424 (2002)... 

In summary, the mathematical similarity of different steady-state solute zones, critical to the success of seemingly unrelated separation methods, demonstrates the impressive unifying power of the basic transport approach to chemical separations. This unity is emphasized again in the next chapter, where we delve into the classification and comparison of separation methods. 

We have recently undertaken a reconsideration of cellulase sequences by analyzing each domain of the sequence separately. Here we describe initial observations and methodologies we are developing for the analysis and comparison of cellulase domains. Novel methods of classification will give greater insight into theoretical, evolutionary and practical approaches to the analysis of cellulase. 

Laboratory analysis of human samples for trace contaminants or their metabolites inevitably produces results that deviate quantitatively from the actual concentrations. Such deviations can, for example, complicate exposure classifications in epidemiologic studies, detection of time trends in exposure, and comparison of studies that use data produced with different analytic methods. Individual laboratories can use standard QA-QC methods to minimize and define the magnitude of the variations. However, federal agencies and statutes, such as CDC, the National Institute of Standards and Technology, and statutes such as CLIA, could play important roles in improving the overall quality of biomonitoring laboratory data and their utility in health-related applications. 

Morales SE, HolbenWE (2011) Linking bacterial identities and ecosystem processes can omic analyses be more than the sum of their parts FFMS Microbiol Ecol 75 2-16 Mulder N, Apweiler R (2007) InterPro and InterProScan tools for protein sequence classification and comparison. Method Mol Biol 396 59-70... 

LeseUier E, West C. Description and comparison of chromatographic tests and che-mometric methods for packed column classification. J Chromatogr A 2007 1158 320-60. 

Phillips, H. 1987. A Comparison of Standard Methods for the Determination of Maximum Experimental Safe Gap (MESG). Proc. International Symposium on the Explosion Hazard Classification of Vapors, Gases, and Dusts, pp. 83-108. Publication NMAB-447. National Materials Advisory Board, Washington, DC. 

H. Van der Voet and P.M. Coenegracht, The evaluation of probabilistic classification methods. Part 2. Comparison of SIMCA, ALLOC, CLASSY and LDA. Anal. Chim. Acta, 209 (1988) 1-27. 

W. Werther, H. Lohninger, F. Stand and K. Vermuza, Classification of mass spectra. A comparison of yes/no classification methods for the recognition of simple structural properties. Chemom. Intell. Lab. Syst., 22 (1994) 63-67. 

FIGURE 5.28 Comparison of the test errors for the glass data using different classification methods. One hundred replications of the evaluation procedure (described in the text) are performed for the optimal parameter choices (if the method depends on the choice of a parameter). The methods are LDA, LR, Gaussian mixture models (Mix), fc-NN classification, classification trees (Tree), ANN, and SVMs. 

---