Standard classification algorithms are generally designed to maximize the number of correct predictions (concordance). The criterion of maximizing the concordance may not be appropriate in certain applications. In practice, some applications may emphasize high sensitivity (e.g., clinical diagnostic tests) and others may emphasize high specificity (e.g., epidemiology screening studies). This paper considers effects of the decision threshold on sensitivity, specificity, and concordance for four classification methods: logistic regression, classification tree, Fisher's linear discriminant analysis, and a weighted k-nearest neighbor. We investigated the use of decision threshold adjustment to improve performance of either sensitivity or specificity of a classifier under specific conditions. We conducted a Monte Carlo simulation showing that as the decision threshold increases, the sensitivity decreases and the specificity increases; but, the concordance values in an interval around the maximum concordance are similar. For specified sensitivity and specificity levels, an optimal decision threshold might be determined in an interval around the maximum concordance that meets the specified requirement. Three example data sets were analyzed for illustrations.
Canonical discriminant analysis (CDA) was applied in order to distinguish the water-quality and the sediment-quality parameters from neighboring rivers, and to recognize similarities of water and sediment properties between a lagoon and neighboring rivers. Two set of constructed discriminant functions showed a marked contribution to most of the discriminant variables. In water, the significant parameters - the total nitrogen, algae, dissolved oxygen and total phosphate - were combined as the nutrient effect factor. The recognition capacities of the two discriminant functions were 95.6 and 4.4%, respectively; the Kaoping River showed the most similarities with the water quality in Dapeng Bay; in sediment, the significant parameters porosity, Cd, Cr, Al, and Pb were combined as the heavy metal effect factor. The recognition capacities were 82.6 and 17.4%, respectively, but the sediment properties in these three rivers had no significant similarity with the Dapeng Bay.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.