Comparing Random Forests Regression with Logistic Regression to Determine the Most Important Factors Affecting Congenital Malformations For Newborn Babies in Iraq
Dr.Haitham Yaqub Yousif, Arshad Hameed Hassan, Shaima Mohammed Ahmed
Congenital malformation is a structural defect in one or more parts of the body from birth , The causes and sources of birth defects may be genetic or caused by a non-genetic event before birth , Some congenital malformations may be caused by taking drugs or sometimes the causes are unknown. In recent years, the rate of congenital malformations among newborn children in Iraq has increased and to identify this problem and identify the most important factors affecting it. A sample of children with congenital malformations was taken to the maternity hospitals of the Health Department of Baghdad / Rusafa and Karkh - Department of preterm infants of 2504 births , To identify the most important factors affecting the congenital malformations using artificial intelligence techniques and machine learning including random forests regression and logistic regression decline as these techniques are one of the most advanced techniques used in the case of big data , We concluded from this research, which aimed to find the best model for estimating the data of congenital anomalies in Iraq through the use of two types of machine learning models (artificial intelligence) and as these types are regression models at the same time and after estimating each model we compared using the mean squares error criterion and it was the best A model is a random forest model regression.
Volume: Volume 24
Issues: Issue 2
Keywords: Tree Regression, Random Forest, Binary Logistic regression, Congenital malformation, Mean Square Error.