Article
Adaptive Intelligence for Early Thyroid Disease Detection and Classification
Adaptive Intelligence for Early Thyroid Disease Detection and Classification presents an intelligent decision-support system designed to assist in the early identification of thyroid disorders using machine learning techniques. The proposed application provides separate modules for administrators and users, enabling efficient dataset management, model training, and disease prediction through a userfriendly interface. The thyroid dataset contains multiple clinical attributes associated with three diagnostic categories: Normal, Hyperthyroid, and Hypothyroid. Before model development, the data undergoes preprocessing, including cleaning, label encoding, and feature transformation to ensure compatibility with machine learning algorithms. Several classification models, including Decision Tree, Random Forest, Support Vector Machine, K-Nearest Neighbors, Logistic Regression, Naïve Bayes, and Multi-Layer Perceptron, are trained and evaluated using performance measures such as accuracy, precision, recall, and F1-score. Comparative analysis identifies Random Forest and Decision Tree as the most effective models for thyroid disease classification. The system also allows users to predict thyroid conditions by entering clinical values or uploading test data. Developed using Python, Django, and MySQL, the proposed framework simplifies disease assessment, supports early diagnosis, reduces manual effort, and provides a reliable platform for intelligent thyroid disease detection.
Full Text Attachment
