Using the discriminant function to determine the factors affecting the weights of New-born’s

Abstract

The normal growth of a child is a good indicator of the child's overall health. If any growth problems occur، it is a sign of underlying issues that need to be addressed. If the child's weight is below normal، they are more susceptible to oxygen deficiency، low blood sugar، and irregular body temperature .However، if the child's weight is above normal، they are more prone to hypoglycaemia and problems for the mother and fetus during childbirth. The research focuses on studying the classification of the child's weight، as most studies rely on the level of injury in children or blood sugar levels .In this research، the method of multivariate statistical analysis was used، specifically the discriminant analysis method. It is considered one of the most important multivariate statistical methods used in processing descriptive data .It relies on constructing a function called the discriminant function، which is a linear combination of a set of independent variables. This function works to reduce the similarity in classification errors .The goal of discriminant analysis is to classify observations into their correct groups with the least possible classification error.It is considered one of the most important multivariate statistical methods used in processing descriptive data. It relies on constructing a function called the discriminant function، which is a linear combination of a set of independent variables. This function works to reduce the similarity in classification errors. The goal of discriminant analysis is to classify observations into their correct groups with the least possible classification error. The research aims to build a model that can achieve a modern classification by formulating a linear discriminant function based on common indicators، determining the importance of these indicators in classification، and testing the accuracy of the classification .The research aims to build a model that can achieve modern classification by formulating a linear discriminant function based on common indicators، understanding the importance of these indicators in classification، and testing the accuracy of the classification .Ninety samples were used، obtained from the City of Medicine Hospital in Baghdad and Al-Batool Hospital in Diyala Governorate، specifically for new-borns. The study concluded after analysis that the predicted correct classification rate in the discriminant function was 63.3%. Out of 613 samples classified into the second group، which originally belonged to the first group، 37 samples were classified correctly، resulting in a correct classification rate of 64.9% with 24 samples classified correctly. As for the classification of samples into the second group، the correct classification rate was 62.3% with 33 samples out of 53 classified correctly، meaning that 20 samples were incorrectly classified into the first group while they actually belonged to the second group. This occurs due to the types of errors discussed in the theoretical section.