Analyzing Data by Applying Neural Networks to Identify Patterns in the Data

This paper analyzes the dataset to identify the patterns in the dataset. Fetal health classification dataset has been taken for analysis. The study touches the fields of medicine such as obstetrics and mortality in it. Currently, there is high maternal and fetal mortality at birth due to limited resources, which could have been prevented. One of the most accessible and simple means of assessing fetal health is the cardiogram (CGT). It can be used to make a diagnosis while still in the womb and prevent death. A dataset containing fetal CGT data was taken as the dataset under study. It consists of 521 records and 23 attributes, among which are: initial fetal heart rate; number of accelerations per second; number of fetal movements per second; number of uterine contractions per second; number of LD per second; number of SD per second; number of PD per second; percentage of time with abnormal short-term variability; mean value of short-term variability; percentage of time with abnormal long-term variability; mean value of long-term variability; width of the histogram constructed using all the values of the histogram; and the width of the histogram constructed using all the values of the histogram; maximum histogram value; number of peaks in the study histogram; number of zeros in the histogram; Hist mode; Hist mean; Hist variance; Hist trend; fetal condition: 1 - normal 2 - suspicious 3 - pathologic; information attribute - patient ID. Data analysis is performed using decision tree, Kohonen maps and neural networks.