Machine Learning Empowered Osteoporosis Prediction: A Comparative Analysis

Osteoporosis is an illness of the bones. Numerous studies have been conducted on the multi-factorial condition known as osteoporosis and fracture of bone as its clinical outcome. A bone condition called osteoporosis arises when bone mass and mineral density decline or when alterations occur to the composition and integrity of bone. In this research work, diverse machine learning algorithms like Logistic Regression (LoR), Gradient Boosting Classifier (GB), Decision Tree Classifier (DT), Naive Bayes (NB), K Neighbors Classifier(KNN), Ada Boost Classifier(AB), Random Forest Classifier(RF), Support Vector Machine Classification(SVC), Voting Classifier(VC) and Stacking Classifier(SC) have been used to predict osteoporosis. Various metrics have been employed such as Accuracy, MCC, F1- Score, FNR., NPV, Sensitivity, Specificity, Precision and FDR. The performance of the Gradient Boosting Algorithm was found to be the best with a testing accuracy of 0.9284 and it can be used to predict Osteoporosis with reliability.