Wyniki wyszukiwania - BazTech

1

Tworzenie reguł diagnostycznych na podstawie danych

Straszecka E.

Pomiary Automatyka Kontrola

|

2014

|

R. 60, nr 4

219--222

PL

W pracy przedstawiono metodę tworzenia reguł diagnostycznych o rozmytych przesłankach reprezentujących objawy i nierozmytej konkluzji odpowiadającej diagnozie. Reguły tworzy się na podstawie danych uczących, lecz są one zrozumiałe dla ekspertów i mogą być przez nich weryfikowane. Zbiór reguł dla każdej z diagnoz jest ustalany odrębnie, z zastosowaniem oryginalnego algorytmu eliminacji reguł. Obliczenia dla dwóch benchmarkowych baz danych potwierdzają efektywność proponowanych metod.

EN

A method of diagnostic rule creation is presented in the paper. The rules have fuzzy premises that represent symptoms and a crisp conclusion relevant to the diagnosis. Each rule has an assigned weight that is determined as a value of the basic probability assignment defined in the Dempster-Shafer theory. Having created the rules, there is performed the diagnostic reasoning for a consulted case whose outcomes are values of the Bel belief measure (of the Dempster-Shafer theory) for all diagnostic hypotheses. The hypothesis of the maximal belief is the ultimate conclusion. Membership functions of symptoms and the basic probability assignment are found from the training data. Although the rules are created by means of data, they are understandable for human experts who can interpret and verified them. An individual set of rules is provided for each diagnosis. It results from an original elimination algorithm that is proposed in the paper. The elimination process starts from the complete set of rules and the algorithm indicates rule(s) of the lowest diagnostic significance, which are next deleted. Numerical experiments for two benchmark databases show the properties of the proposed method.

2

The basic probability assignment as a measure of diagnostic rules significance

Straszecka E.

Journal of Medical Informatics & Technologies

|

2013

|

Vol. 22

95--102

EN

Diagnostic rules are usually IF-THEN rules, but they should satisfy specific requirements of a diagnosis. Thus, not always the classical methods of rules determination are applicable. In the present paper it is suggested to find out the set of rules by an elimination of superfluous rules from the maximal rule set or adding rules that improve inference to the minimal set of rules. It is shown that the basic probability assignment determined in the Dempster-Shafer theory of evidence can be used as a measure indicating symptoms that are the most significant for a diagnosis and should create rules. A set of IF-THEN rules with fuzzy premises and crisp conclusions can be built in this way. The proposed method is illustrated by determining rules allowing for diagnostic inference for a database of thyroid gland diseases.

3

A new approach to multiple fault diagnosis: A combination of diagnostic matrices, graphs, algebraic and rule-based models. The case of two-layer models

Ligęza A., Kościelny J. M.

International Journal of Applied Mathematics and Computer Science

|

2008

|

Vol. 18, no 4

465-476

EN

The diagnosis of multiple faults is significantly more difficult than singular fault diagnosis. However, in realistic industrial systems the possibility of simultaneous occurrence of multiple faults must be taken into account. This paper investigates some of the limitations of the diagnostic model based on the simple binary diagnostic matrix in the case of multiple faults. Several possible interpretations of the diagnostic matrix with rule-based systems are provided and analyzed. A proposal of an extension of the basic, single-level model based on diagnostic matrices to a two-level one, founded on causal analysis and incorporating an OR and an AND matrix is put forward. An approach to the diagnosis of multiple faults based on inconsistency analysis is outlined, and a refinement procedure using a qualitative model of dependencies among system variables is sketched out.