Model uncertainty and bias in the evaluation of nuclear spectra
- 27 Downloads
Unbiased analysis of γ-ray and X-ray spectra is impossible in the absence of a complete physical or mathematical model. Partial model knowledge may be supplemented by simple assumptions or by various heuristic schemes in order to effect a solution. Assessment of limits for bias, based upon the properties of the surrogate model and physical-chemical knowledge of the measurement system, is the principal target of this work. The Smoothest Consistent Baseline (SCB) approach has been introduced in an attempt to reduce assumptions and minimize bias in the extraction of a spectral peak from a baseline of uncertain shape. The bias matrix, which results directly from the numerical analysis, permits limiting baseline profiles to be simply converted into bounds for systematic model error.
KeywordsModel Uncertainty Counting Statistic True Baseline Bias Vector Linear Baseline
Unable to display preview. Download preview PDF.
- 1.J. OP DE BEECK, At. Energy Rev., 13 (1975) 743.Google Scholar
- 2.P. QUITTNER, Gamma-Ray Spectroscopy—with Special References to Computer and Detector Techniques, Hilger, London, Akadémiai Kiadó, Budapest, 1972.Google Scholar
- 6.L. A. CURRIE, in Modern Trends in Activation Analysis, J. R. DeVOE, P. D. LaFLEUR (Eds), Vol. 2 US NBS Spec. Publ. 312, 1969, p. 1215.Google Scholar
- 7.L. A. CURRIE, Anal. Letters, 4 (1971) 873.Google Scholar
- 8.H. P. YULE, in Modern Trends in Activation Analysis J. R. DeVOE, P. D. LaFLEUR (Eds), Vol. 2 US NBS Spec. Publ. 312, 1969, p. 1155.Google Scholar
- 14.R. GUNNICK, J. B. NIDAY, Rep. UCRL-51061, 1, 1972.Google Scholar
- 15.F. H. SCHAMBER, in X-Ray Fluorescence Analysis of Environmental samples, Chapter 21, T. DZUBAY (Ed.), Ann Arbor Science Publ., Ann Arbor, 1976.Google Scholar
- 16.C. S. G. BEVERIDGE, R. S. SCHECHTER, Optimization: Theory and Practice, Chapter 7, McGraw-Hill, New York, 1970.Google Scholar