Thyroid Disease Prediction Using Selective Features and Machine Learning Techniques

Abstract

Thyroid disease prediction has emerged as an important task recently. Despite existing approaches for its diagnosis, often the target is binary classification, the used datasets are small-sized and results are not validated either. Predominantly, existing approaches focus on model optimization and the feature engineering part is less investigated. To overcome these limitations, this study presents an approach that investigates feature engineering for machine learning and deep learning models. Forward feature selection, backward feature elimination, bidirectional feature elimination, and machine learning-based feature selection using extra tree classifiers are adopted. The proposed approach can predict Hashimoto’s thyroiditis (primary hypothyroid), binding protein (increased binding protein), autoimmune thyroiditis (compensated hypothyroid), and non-thyroidal syndrome (NTIS) (concurrent non-thyroidal illness). Extensive experiments show that the extra tree classifier-based selected feature yields the best results with 0.99 accuracy and an F1 score when used with the random forest classifier. Results suggest that the machine learning models are a better choice for thyroid disease detection regarding the provided accuracy and the computational complexity. K-fold cross-validation and performance comparison with existing studies corroborate the superior performance of the proposed approach.

A comparative study of thyroid data classification based on GA, BPSO, and ACO metaheuristics approaches

“…In our case, the algorithm was integrated with three metaheuristic algorithms as a classifier in the selection of the most important parameters of the model. Metaheuristic techniques prefer to integrate classifiers, and according to these studies, this algorithm has provided satisfactory results 7,16 . It is a classifier that uses several decision trees in various subgroups of a dataset and generally improves the forecast accuracy.…”

“…Meanwhile, in 7 , a method that uses feature selection and combines deep learning and machine learning models has been suggested. The selected characteristics of the RF classifier provide an accuracy of 99%.…”

Efficient Feature Extraction Based on Optimized Gated Recurrent Unit Recurrent Neural Network for Automatic Thyroid Prediction

“…An accuracy improvement of 0.7% was made using the 6-entity MMLP in comparison to a stand-alone system. Rajasekhar Chaganti et al, [14] have introduced a method of examining feature design for AI and deep learning models. Includes robustness employing extra AI-based tree classi ers, bilateral feature decision, forward feature selection, inverse feature removal, and robustness.…”

Machine Learning-Based Prediction of Hashimoto’s Thyroiditis Development Risk

“…In recent years, there has been a growing interest in using ML for the diagnosis and prognosis of HT. 8,9 This study aims to develop a ML model to predict the risk of HT development with high accuracy. To achieve this, different ML algorithms will be employed, and their hyperparameters will be optimized.…”

“…In recent years, there has been a growing interest in using ML for the diagnosis and prognosis of HT. 8,9…”