Succinct Differentiation of Disparate Boosting Ensemble Learning Methods for Prognostication of Polycystic Ovary Syndrome Diagnosis

TL;DR

This paper compares different boosting ensemble methods for diagnosing Polycystic Ovary Syndrome using clinical data, highlighting their performance differences and data anomalies to improve prognostic accuracy.

Contribution

It provides a detailed differentiation and performance analysis of Adaptive Boost, Gradient Boosting, XGBoost, and CatBoost for PCOS diagnosis.

Findings

XGBoost and CatBoost outperform other methods in accuracy

Data anomalies significantly affect model performance

Performance metrics vary across different boosting techniques

Abstract

Prognostication of medical problems using the clinical data by leveraging the Machine Learning techniques with stellar precision is one of the most important real world challenges at the present time. Considering the medical problem of Polycystic Ovary Syndrome also known as PCOS is an emerging problem in women aged from 15 to 49. Diagnosing this disorder by using various Boosting Ensemble Methods is something we have presented in this paper. A detailed and compendious differentiation between Adaptive Boost, Gradient Boosting Machine, XGBoost and CatBoost with their respective performance metrics highlighting the hidden anomalies in the data and its effects on the result is something we have presented in this paper. Metrics like Confusion Matrix, Precision, Recall, F1 Score, FPR, RoC Curve and AUC have been used in this paper.