Scholarlink

<p>The complex and heterogeneous clinical presentation of Systemic Lupus Erythematosus, a chronic autoimmune disease, makes diagnosis and treatment extremely difficult. Traditional methods often fail to capture the intricate patterns present in lupus datasets, underlining the possibility of advanced computational approaches such as Machine Learning and Quantum Machine Learning. In this study, the performance of classical machine learning models, such as the Support Vector Machine (SVM), Random Forest (RF) and Neural Networks (NN), is evaluated and compared to the performance of their quantum equivalents, which are Quantum Support Vector Machine (QSVM), Quantum Random Forest (QRF), and Quantum Neural Networks (QNN). To evaluate the efficacy of these models in determining the remission, they were put through a series of tests using two sets of features, i.e., 10 features and 20 features, spread over two remission states, namely Remission 1 and Remission 1-5. Experiments show how feature dimensionality and remission state affect model achievement, which helps us understand the relative strengths of classical and quantum approaches. The results showed how both types of models remain relevant in the case of the study, with the highest score of 88% and 79% accuracy for both experiments, reached with RF and QNN, respectively.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Integrating quantum and classical machine learning for predicting remission in systemic lupus erythematosus

  • 期刊论文

  • 发表日期：

    2025年11月10日

查看全文
  • Pablo A. Osorio-Marulanda,
  • Ubaid Ullah,
  • Guillermo Ruiz-Irastorza,
  • Begoña García-Zapirain

摘要

The complex and heterogeneous clinical presentation of Systemic Lupus Erythematosus, a chronic autoimmune disease, makes diagnosis and treatment extremely difficult. Traditional methods often fail to capture the intricate patterns present in lupus datasets, underlining the possibility of advanced computational approaches such as Machine Learning and Quantum Machine Learning. In this study, the performance of classical machine learning models, such as the Support Vector Machine (SVM), Random Forest (RF) and Neural Networks (NN), is evaluated and compared to the performance of their quantum equivalents, which are Quantum Support Vector Machine (QSVM), Quantum Random Forest (QRF), and Quantum Neural Networks (QNN). To evaluate the efficacy of these models in determining the remission, they were put through a series of tests using two sets of features, i.e., 10 features and 20 features, spread over two remission states, namely Remission 1 and Remission 1-5. Experiments show how feature dimensionality and remission state affect model achievement, which helps us understand the relative strengths of classical and quantum approaches. The results showed how both types of models remain relevant in the case of the study, with the highest score of 88% and 79% accuracy for both experiments, reached with RF and QNN, respectively.