Enhanced Accuracy in Thyroid Disease Classification: A Comparative Analysis of Random Forest and Decision Tree Methods

Abstract

In rural areas where rapid diagnosis of lifestyle conditions is often unavailable, the development of intelligent prediction systems using modern computing techniques is imperative. This study aimed to enhance the accuracy of thyroid disorder identification by creating a more precise classification model. The study compared the performance of two machine learning techniques, namely Random Forest and Decision Tree, to determine the optimal training model for detecting thyroid disorders. Utilizing datasets from the UCI machine learning library, these classifiers were applied to differentiate between patients with hyperthyroidism and hypothyroidism. Performance measurements such as Accuracy and Precision were used to assess the models. The Decision Tree classifier achieved an 84 percent accuracy rate, while the Random Forest classifier demonstrated an 85 percent accuracy rate. Similarly, the precision rates for the Decision Tree and Random Forest models were calculated to be 82 percent and 84 percent, respectively. These findings suggest that both classifiers offer promising results in accurately identifying thyroid disorders, with the Random Forest model exhibiting slightly higher accuracy and precision.