استخدام التحليل التمييزي الخطي لتصنيف مرضى الفشل الكلوي

Abstract

This research aimed at classifying renal failure diseases whether acute renal failure (ARF) or chronic renal failure (CRF) based on linear discriminant analysis (LDA). The research was carried out by collecting data from patients with renal failure at Tobruk Medical Center (TMC). The sample was composed of 461 cases of renal failure, 238 cases of them were acute renal failure and 223 cases were chronic renal failure taking into account a set of crucial factors, including age, fasting blood sugar, urine, etc. The discriminant function analysis was applied using the R programming language. The study revealed that the linear discriminant analysis classified the two types and indicated that the variables with the highest impact on renal failure were creatinine and fat blood sugar. Also, the LDA was highly precise in classification, where 88.5% of the sample was classified successfully, which means a small error rate of 11.5%. Keywords: linear Discriminant Analysis, renal failure, classification.