Affected States soft independent modeling by class analogy from the relation between independent variables, number of independent variables and sample size.

Objective: The aim of study is to introduce method of Soft Independent Modeling of Class Analogy (SIMCA), and to express whether the method is affected from the number of independent variables, the relationship between variables and sample size. Study Design: Simulation study. Material and Methods: SIMCA model is performed in two stages. In order to determine whether the method is influenced by the number of independent variables, the relationship between variables and sample size, simulations were done. Conditions in which sample sizes in both groups are equal, and where there are 30, 100 and 1000 samples; where the number of variables is 2, 3, 5, 10, 50 and 100; moreover where the relationship between variables are quite high, in medium level and quite low were mentioned. Results: Average classification accuracy of simulation results which were carried out 1000 times for each possible condition of trial plan were given as tables. Conclusion: It is seen that diagnostic accuracy results increase as the number of independent variables increase. SIMCA method is a method in which the relationship between variables are quite high, the number of independent variables are many in number and where there are outlier values in the data that can be used in conditions having outlier values. Turkish Baslik: Analojik Siniflamada Esnek Bagimsiz Modelinin (ASEBAM), Bagimsiz Degiskenler Arasindaki Iliski, Bagimsiz Degisken Sayisi ve Orneklem Buyuklugunden Etkilenme Durumu Anahtar Kelimeler: Siniflama, Coklu bagimlilik, Asiri uc deger Amac: Calismanin amaci, Analojik Siniflamada Esnek Bagimsiz Model (ASEBAM) yontemi tanitmak, yontemin bagimsiz degisken sayisi, degiskenler arasindaki iliski durumu ve orneklem buyuklugunden etkilenip etkilenmedigini ortaya koymaktir. Gerec ve Yontemler: ASEBAM modeli iki asamada gerceklestirilmektedir. Yontemin bagimsiz degisken sayisi, degiskenler arasindaki iliski ve orneklem buyuklugunden etkilenip etkilenmedigini ortaya koymak amaci ile simulasyon denemeleri yapilmistir. Her iki gruptaki orneklem buyukluklerinin esit ve 30, 100 ve 1000 oldugu, degisken sayisinin 2, 3, 5, 10, 50 ve 100 oldugu durumlar, ayrica degiskenler arasindaki iliskilerin cok yuksek (0.95), orta duzeyde (0.50) ve cok dusuk (0.05) oldugu durumlar dikkate alinmistir. Her bir kombinasyon 1000 kez denenmistir. Bulgular: Deneme planina ait her bir olasi durum icin 1000 kez gerceklestirilen simulasyon sonuclarinin ortalama siniflama dogruluklari tablo halinde verilmistir. Sonuc: Bagimsiz degisken sayisi artikca diagnostik dogruluk sonuclarinin artigi gorulmektedir. ASEBAM metodu degiskenler arasinda iliskilerin cok yuksek, bagimsiz degisken sayisinin cok fazla ve veride asiri uc degerlerin var oldugu durumda da kullanilabilecek istatistik anlamlilik degeri var olan bir yontemdir.