Detecting Thyroid Disease Using Optimized Machine Learning Model Based on Differential Evolution

Abstract

Thyroid disease has been on the rise during the past few years. Owing to its importance in metabolism, early detection of thyroid disease is a task of critical importance. Despite several existing works on thyroid disease detection, the problem of class imbalance is not investigated very well. In addition, existing studies predominantly focus on the binary-class problem. This study aims to solve these issues by the proposed approach where ten types of thyroid diseases are considered. The proposed approach uses a differential evolution (DE)-based optimization algorithm to fine-tune the parameters of machine learning models. Moreover, conditional generative adversarial networks are used for data augmentation. Several sets of experiments are carried out to analyze the performance of the proposed approach with and without model optimization. Results suggest that a 0.998 accuracy score can be obtained using AdaBoost with DE optimization which is better than existing state-of-the-art models.

Machine Learning Approach with Harmonized Multinational Datasets for Enhanced Prediction of Hypothyroidism in Patients with Type 2 Diabetes

“…Some of these approaches have included ensemble learning [ 43 ] and the implementation of Local Interpretable Model-Agnostic Explanations and prediction results within smartphone applications for easy-to-interpret results [ 89 ]. In regard to HT prediction using ML, studies have worked to overcome diagnostic challenges that pertain to equivocal symptoms [ 43 ], and challenges related to the data itself, such as class imbalances [ 90 ]. This was achieved, respectively, by elucidating important symptoms that contribute to HT and non-HT classification [ 43 ] and generating artificial data to include model development and testing using a Conditional Tabular Generative Adversarial Network [ 90 ].…”

“…In regard to HT prediction using ML, studies have worked to overcome diagnostic challenges that pertain to equivocal symptoms [ 43 ], and challenges related to the data itself, such as class imbalances [ 90 ]. This was achieved, respectively, by elucidating important symptoms that contribute to HT and non-HT classification [ 43 ] and generating artificial data to include model development and testing using a Conditional Tabular Generative Adversarial Network [ 90 ]. Similar to our use of distinct datasets to test how diversified data impact model performance, Hu et al used patient data from four clinical sites to examine the generalizability of an ML to predict thyroid disease; the model performed well, with an AUROC of 90.9%, and the external validity was confirmed by the lack of bias within the different sites [ 91 ].…”

Innovative Framework for Thyroid Disease Detection by Leveraging Hybrid AGTEO Feature Selection and GRU Classification Model

“…Punit Gupta et al, proposed an approach which uses a differential evolution (DE)-based optimization algorithm to fine-tune the parameters of machine learning models [17] [18]. The main contributions were an enhanced feature selection accuracy of the dataset and a multiclass classification to differentiate flanked by three different types of thyroid disorders.…”

“…However, there is a lack of research on evaluating these models on different datasets other than the UCI repository data [16]. However, while Gupta et al Vgg-19-LSTM has the best performance of all the LSPs, there is still a gap in research concerning whether it is generalizable to different datasets, and whether it is robust to changes in the quality of images [17]. Alnaggar et al present a XGBoost-based classifier with good accuracy, but the research shortfall comes in evaluating the model's performance on different datasets, as well as testing the model's scalability [18].…”