Scholarlink

<p>Age-based disease prediction and severity assessment are significant challenges in personalized healthcare, often limited by a lack of model interpretability. In this study, we propose an integrated approach combining explainable AI techniques Shapley Additive explanations (SHAP) and local interpretable model-agnostic explanations (LIME) with deep learning methods for more accurate and interpretable disease prediction and severity assessment. We evaluate six classification algorithms-Support Vector Machine, Naive Bayes, Random Forest, Decision Tree, and Logistic Regression-on age-based datasets, finding that Random Forest achieves the highest accuracy. To enhance interpretability, we apply SHAP and LIME to the Random Forest model, offering insights into feature importance and individual predictions. Additionally, we introduce a long short-term memory (LSTM) network to predict disease severity, leveraging age, habits, and symptoms as input features. The LSTM model demonstrates accurate severity prediction through rigorous training and testing. We also developed an interface that allows healthcare professionals to input patient data and generate a severity prediction graph for informed decision-making. The main contributions of this paper include the integration of explainable AI with deep learning for more transparent disease prediction, the introduction of an LSTM-based model for disease severity, and the development of a practical decision support tool for healthcare providers. Experimental results demonstrate that Random Forest achieves superior classification accuracy, while the LSTM model reliably predicts disease severity, offering valuable insights for personalized healthcare interventions.</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

Age-based disease prediction and health monitoring: integrating explainable AI and deep learning techniques

  • 期刊论文

  • 发表日期：

    2025年1月9日

查看全文
  • G. L. N. D. Sushmitha,
  • Sairam Utukuru

摘要

Age-based disease prediction and severity assessment are significant challenges in personalized healthcare, often limited by a lack of model interpretability. In this study, we propose an integrated approach combining explainable AI techniques Shapley Additive explanations (SHAP) and local interpretable model-agnostic explanations (LIME) with deep learning methods for more accurate and interpretable disease prediction and severity assessment. We evaluate six classification algorithms-Support Vector Machine, Naive Bayes, Random Forest, Decision Tree, and Logistic Regression-on age-based datasets, finding that Random Forest achieves the highest accuracy. To enhance interpretability, we apply SHAP and LIME to the Random Forest model, offering insights into feature importance and individual predictions. Additionally, we introduce a long short-term memory (LSTM) network to predict disease severity, leveraging age, habits, and symptoms as input features. The LSTM model demonstrates accurate severity prediction through rigorous training and testing. We also developed an interface that allows healthcare professionals to input patient data and generate a severity prediction graph for informed decision-making. The main contributions of this paper include the integration of explainable AI with deep learning for more transparent disease prediction, the introduction of an LSTM-based model for disease severity, and the development of a practical decision support tool for healthcare providers. Experimental results demonstrate that Random Forest achieves superior classification accuracy, while the LSTM model reliably predicts disease severity, offering valuable insights for personalized healthcare interventions.