Wyniki wyszukiwania - Biblioteka Nauki

Nowa wersja platformy, zawierająca wyłącznie zasoby pełnotekstowe, jest już dostępna.
Przejdź na https://bibliotekanauki.pl

Ograniczanie wyników

1 2010

Znaleziono wyników: 1

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: probabilistyczno-rozmyta baza wiedzy

Sortuj według:

Ogranicz wyniki do:

Rozmyty system wnioskujący o modelu bazującym na regułach asocjacji

100%

Rudnik K. , Walaszek-Babiszewska A.

tom Vol. 13, nr 2

50-60

In the practice of company management we deal with many tasks, that are associated with limited knowledge and uncertainty about the course of events and activities managed objects. Fuzzy IF-THEN rules are an appropriate form of describing subjective uncertainty results from lack of knowledge and objective uncertainty results from characteristics of different processes. On the other hand, the uncertainty due to randomness can be described using the theory of probability. The paper presents inference system with probabilistic-fuzzy knowledge base as a tool which can help user to analyze complete uncertainty of real problems in the company using fuzzy sets and probability. In the mentioned system, knowledge is saved in the weighted IF-THEN fuzzy rules, where the weights constitute marginal probabilities of the fuzzy events in the antecedents and conditional probabilities of the fuzzy events in the consequents. Moreover, this paper propose using fuzzy association rules as a method of automatic knowledge base extraction in the inference system. For this purpose a modification of the Apriori algorithm was described. The algorithm extracts the most important and matching linguistic rules by assumption of minimum support as a minimum joint probability of the events in the rules. If minimum support equals zero, then the rules present total probabilistic distribution of the fuzzy events, otherwise the rules present probabilistic distribution, which is the best matching to a variables universe. In the methodology of system creation, the universe of quantitative variables is discretized on disjoint intervals of variable values and the fuzzy sets are defined by grades of membership of the disjoint intervals to fuzzy sets. This approach allows vectorize the calculation. A numerical example is analyzed by using a wind speed prediction process. Parameter of wind speed characterized by high variability of random character. However, the correct estimation of wind speed, as a energy resources, is necessary for control working of wind turbine. It is also important for the localization process of wind turbines, production planning and estimating cost-effectiveness of such investments.