Estimating the Costs Associated with Worthwhile Predictions of Poor Air Quality
In this study we investigate the effect of varying the ratio of false-positive and false-negative misclassification costs on the sensitivity and selectivity of binary predictions of exceedences of atmospheric pollutants. This allows us to determine a window of values far this ratio far which it is worthwhile making definite rather than probabilistic predictions. The support vector machine provides a suitable statistical pattern recognition method for this work.
KeywordsSupport Vector Machine Training Pattern Misclassification Cost Support Vector Classification Radial Basis Func
Unable to display preview. Download preview PDF.
- B. Boser, I. Guyon, and V. N. Vapnik, “A training algorithm for optimal margin classifiers” in Proceedings of the fifth annual workshop on computational learning theory, (Pittsburgh), pp. 144–152, ACM, 1992.Google Scholar
- C. Cortes and V. Vapnik, “Support vector networks” Machine Learning, vol. 20, pp. 1–25, 1995.Google Scholar
- Y. Lin, Y. Lee, and G. Wahba, “Support vector machines for classification in nonstandard situations” Tech. Rep. 1016, Department of Statistics, University of Wisconsin, 1210 West Drayton St., Madison, WI 53’706, March 2000.Google Scholar
- K. Veropoulos, C. Campbell, and N. Cristianini, “Controlling the sensitivity of support vector machines” in Proceedings of the International Joint Conference on Artificial Intelligence (Workshop ML3), (Stockholm, Sweeden), pp. 17–21, 1999.Google Scholar
- G. C. Cawley, “MATLAB SVM toolbox (vO.50) [http://theoval.sys.uea.ac.uk/-gcc/svm /toolbox].” University of East Anglia, School of Information Systems, Norwich, Norfolk, UK NR4 7TJ, 2000.Google Scholar
- G. C. Cawley, “Model selection for support vector machines via adaptive step-size tabu search” in Proceedings of the International Conference on Artificial Neural Networks and Genetic Algorithms (accepted for publication), (Prague), April 2001.Google Scholar
- T. Joachims, “Estimating the generalization performance of a SVM efficiently” Tech. Rep. LS-8 number 25, Univerität Dortmund, Fachbereich Informatik, 1999.Google Scholar