Predictive Analytics on Female Infertility Using Ensemble Methods

Abstract

With the accessibility of healthcare data for a significant proportion of patients in hospitals, using predictive analytics to detect diseases earlier has become more feasible. Identifying and recording key variables that contribute to a specific medical condition is one of the most difficult challenges for early detection and timely treatment of diseases. Conditions such as infertility that are difficult to detect or diagnose can now be diagnosed with greater accuracy with the help of predictive modeling. Infertility detection, particularly in females, has recently gained attention. In this work, the researchers proposed an intelligent prediction for female infertility (PreFI). The researchers use 26 variables for the early diagnosis and determine a subset of these 26 variables as biomarkers. These biomarkers contribute significantly to a better prediction of the problem. The researchers designed PreFI using ensemble methods with biomarkers and improved the performance of the predictive system.