TUBERKULOZ HASTALI?ININ TE?HY?Y YÇYN LOJYK SADELE?TYRME METODU YLE BYR UZMAN SYSTEM TASARIMI

Abstract

Bu çaly?mada; TB hastaly?ynyn te?hisi için kontrollü bir uzman sistem (US) tasarlanmy? ve TB hastaly?ynyn belirtileri (12 belirti, 2^12=4096 de?i?ik durum) dikkate alynarak do?ruluk tablosu olu?turulmu?tur. TB hastaly?ynyn belirtilerine göre 6 farkly durum çyky? de?eri olarak alynmy?tyr ve indirgenmi? kural tabanlary elde edilmi?tir. Elde edilen bu de?erler bir uzman sistemde kullanylmy? ve uzman sistemler yardymyyla TB hastaly?y te?his edilmeye çaly?ylmy?tyr. Sonuç olarak bu çaly?mada, US ile yapylan TB te?hisinde iyi sonuçlaryn alyndy?y gözlemlenmi?tir. Yapylan analizler sonucunda US ile yapylan kontrolün hem daha az zaman, daha az ihtimal, hemde gerçek ve güvenilir te?hisler sa?lady?y ve bunun gerçek hayatta kullanylabilirli?i görülmü?tür.

Keywords

• Lojik sadele?tirme, indirgenmi? kural tabanlary, Uzman Sistem, Minimumla?tyrma Metodu, TB te?hisi,

DESIGNING AN EXPERT SYSTEM FOR DETECTION OF TUBERCULOSIS DISEASE WITH LOGIC SIMPLIFICATION METHOD

Öz

In this study; a controlled Expert System (ES) have designed to diagnosis of Tuberculosis (TB) disease and truth table have created by considering the probabilities of TB (12 symptom, 212=4096 different cases). According to the probabilities of TB disease, 6 different cases have accepted as output values and reduced rule bases have obtained. These output values have been processed by an ES and have tried to diagnosis of the TB disease with help to ES. As a result; in this study, we have obtained very good results in diagnosis of the TB with ES. The results of analyses have indicated that the controls performed with ES have provided both less time, less probability and reliable, consistent diagnosis and that are feasible to in real life.

Anahtar Kelimeler

• Logic Simplification, reduced rule based, Expert System, Minimization Method, diagnosis of TB,