Possibilistic Representations of Measurement Combination Problems

This proposal is to do basic research on the use of possibility theory to represent uncertainty in measurement combination problems. Possibility theory is a new form of information theory based on both one-point distribution functions of consistent random sets and a generalization of interval analysis. Possibility theory is closely related to fuzzy sets, fuzzy measures, and Dempster-Shafer evidence theory; however, unlike them, it is based on objective measurement. Possibility distributions provide a natural form for the representation of uncertainty when the measurand or the measuring device is complex, and only small samples from multiple measurement sources are available. Graduate Fellow, Systems Science, SUNY--Binghamton, 327 Spring St. # 2, Portland ME, 04102, USA, (207) 774-0029, cjoslyn@bingsuns.cc.binghamton.edu, joslyn@kong.gsfc.nasa.gov y Supported under NASA Grant # NGT 50756. Joslyn --- Possibilistic Measurement 1 Possibility theory is being developed as an alternativ..