Predicting Thryroid disease using Machine learning methods

Hypothyroidism or hyperthyroidism is a major disease in India which arises due to malfunctioning of thyroid hormones. Medical industry has enormous quantity of data, but the bulk of this data is not processed. For proper diagnosis data must be processed accurately. For accurate processing intelligent Machine Learning techniques are widely used. In this paper an attempt is made to analyze Logistic regression and Support Vector Machine (SVM) for multiclass classification of thyroid dataset. Performance of these techniques is on basis of Precision, Recall, F measure, ROC, RMS Error and accuracy. Our analysis shows that logistic regression is more efficient than SVM for multiclass classification of thyroid dataset.