Sampling and Interpretation of Atmospheric Science Experimental Data
A basic problem in the design of atmospheric experiments is presented by the choice of a sampling rate for the measurement of experimental variables. An approach to the solution of this problem is presented under the assumption that the sampling rate decision can be made prior to the execution of the experiment, as opposed to being made while the experiment is in progress. The technique used is to employ a newly developed and versatile family of fuzzy clustering algorithms, the Fuzzy c-Elliptotypes algorithms, and then to assess the fuzziness of the algorithmically determined clusters as a measure of the quality of the data.
KeywordsCluster Configuration Linear Cluster Probabilistic Uncertainty Cluster Validity Measure Atmospheric Experiment
Unable to display preview. Download preview PDF.
- 1.J. C. Bezdek, R. W. Gunderson, et al., Detection and Characterization of Cluster Substructure, submitted for publication.Google Scholar
- 2.J. C. Bezdek, “Fuzzy Mathematics in Pattern Classification,” Ph.D. Thesis, Cornell University, Ithaca (1973).Google Scholar
- 3.J. C. Bezdek, Mathematical Models for Systematics and Taxonomy, Proc. Eighth Annual Int. Conf. on Numerical Taxonomy, Freeman, San Francisco (1975).Google Scholar
- 5.R. W. Gunderson, Application of Fuzzy ISODATA Algorithms to Star-Tracker Pointing Systems, Proc. 7th Triennial IFAC World Congress, Helsinki (1978).Google Scholar
- 6.J. C. Bezdek and R. W. Gunderson, On the Extension of Fuzzy k-Means Algorithms for Detection of Linear Clusters, Proc. of IEEECDC (Jan. 1979).Google Scholar
- 7.R. V. Duda and P. E. Hart, “Pattern Classification and Scene Analysis,” Wiley-Interscience (1973).Google Scholar
- 8.A. I. Khinchin, “Mathematical Foundations of Information Theory,” Dover (1957).Google Scholar
- 9.L. C. Howlett and K. D. Baker, Development of a rocket-borne reasonance lamp system for measurement of atomic oxygen, Air Force Geophysics Lab., AFGL-TR-77–0227 (Aug. 1977).Google Scholar