Książka = Book ; KS/9/2011/T1P14

Bibliografia s. 173

[1-4], 159-173 stron ; 21 cm

[1-4], 159-173 pages ; 21 cm

Human intelligence is able to solve problems with high amount of uncertainty. Also artificial intelligence tries to solve similar problems. Towards realizing this aim it uses probability theory (PrTh), fuzzy set theory, possibility theory, Dempster-Shafer theory, info-gap theory, etc. PrTh as the oldest one (XVII century) seems to be a ripe and well grounded scientific method. However, according to many opinions, it is not true. In this paper the author shows that the basic and commonly used formula for calculation probability of an event A, p(A) = nA/n, is both qualitatively and quantitatively rather incorrect. This formula was suggested by the frequency interpretation of probability. Furthermore, the author presents the evidential completeness interpretation of probability that seems better suited to describe uncertainty. This interpretation explains why in most cases probability cannot be determined precisely and that only an uncertainty interval of probability can be found.

Bibliography p. 173