İMMÜNOTERAPİ TEDAVİSİNİN BAŞARISINI ETKİLEYEN FAKTÖRLER: BİR İZ ANALİZİ ÇALIŞMASI

Abstract

Valid causal inferences are of great importance in both medical and social research and public policy assessments. In this context, one of the most frequently used methods in interpreting the direct and indirect relationships between the relevant variables is path analysis. In this study, causal relationships between the result, age, gender, type of wart, number of warts, time before treatment, area of the largest wart, and induration diameter were determined by path analysis after estimation of the treatment results of 90 warts patients treated with immunotherapy. After the attribute selection made according to the analysis results, the treatment result estimation application was repeated with 3 most effective variables and the estimations made were found to have the same accuracy and were around 85.2%. The findings showed that the most effective variables on the success of immunotherapy treatment were time to start treatment, age and wart type parameters. In addition, it has been understood that path analysis is a suitable method to determine the most effective independent variables on the result in order to estimate the dependent variable.

Keywords

• wart treatment
• path analysis
• causal graph
• feature selection
• classification