Thyroid Disease Detection Using Machine Learning and Pycaret

Abstract

This work focuses on the analysis and classification models used in the prediction of thyroid disease, using data obtained from the UCI machine learning repository. Machine learning plays a crucial part in the process of disease prediction. In order to better predict the disease based on the parameters obtained from the dataset, this study applies multiple machine learning methods, including decision tree, random forest algorithm, KNN, and Naive Bayes, to the dataset for comparative comparison. The dataset was also modified to improve classification prediction accuracy. The proposed system uses Pycaret to apply various ML algorithms to the dataset, and then compares the results to improve the accuracy with which diseases can be predicted. The Naive Bayes classifier is the most accurate of these, at 95.91 percent.