Scholarlink

<p>Autism Spectrum Disorder (ASD) is a complex neurodevelopmental condition that impacts communication, behavior, and social interaction, affecting individuals across all age groups. Early diagnosis is crucial for effective intervention, typically involving behavioral assessments, genetic testing, and developmental screening. However, the process is often time-consuming and subjective, highlighting the need for automated and objective diagnostic tools. This research explores the integration of Explainable Artificial Intelligence (XAI) with Machine Learning (ML) models to enhance the accuracy and interpretability of Autism Spectrum Disorder diagnosis. Utilizing a diverse dataset comprising 20 attributes and 2,942 patients across different age groups, including toddlers, children, and adults, we applied several Machine Learning algorithms, including Logistic Regression (LR), Decision Trees (DT), Random Forest (RF), Support Vector Machines (SVM), K-Nearest Neighbors (KNN), Convolutional Neural Networks (CNN), and Extreme Gradient Boosting (XGBoost). Extreme Gradient Boosting demonstrated the highest accuracy at 99.05%. Additionally, we employed Explainable AI methods like Local Interpretable Model-agnostic Explanations (LIME) and SHapley Additive exPlanations (SHAP) to provide transparency and deeper insights into the model’s predictions. This framework not only offers reliable, high-accuracy ASD diagnosis but also ensures model interpretability, making it a valuable tool for clinical use and early intervention.</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 explainable AI with machine learning for reliable autism spectrum disorder diagnosis

  • 期刊论文

  • 发表日期：

    2025年11月13日

查看全文
  • Farida Siddiqi Prity,
  • Md. Rehan Ali,
  • Mostafejur Rahman,
  • Mirza Raquib,
  • Md. Maruf Hossain,
  • Md. Mubarak Hossain,
  • K. M. Aslam Uddin

摘要

Autism Spectrum Disorder (ASD) is a complex neurodevelopmental condition that impacts communication, behavior, and social interaction, affecting individuals across all age groups. Early diagnosis is crucial for effective intervention, typically involving behavioral assessments, genetic testing, and developmental screening. However, the process is often time-consuming and subjective, highlighting the need for automated and objective diagnostic tools. This research explores the integration of Explainable Artificial Intelligence (XAI) with Machine Learning (ML) models to enhance the accuracy and interpretability of Autism Spectrum Disorder diagnosis. Utilizing a diverse dataset comprising 20 attributes and 2,942 patients across different age groups, including toddlers, children, and adults, we applied several Machine Learning algorithms, including Logistic Regression (LR), Decision Trees (DT), Random Forest (RF), Support Vector Machines (SVM), K-Nearest Neighbors (KNN), Convolutional Neural Networks (CNN), and Extreme Gradient Boosting (XGBoost). Extreme Gradient Boosting demonstrated the highest accuracy at 99.05%. Additionally, we employed Explainable AI methods like Local Interpretable Model-agnostic Explanations (LIME) and SHapley Additive exPlanations (SHAP) to provide transparency and deeper insights into the model’s predictions. This framework not only offers reliable, high-accuracy ASD diagnosis but also ensures model interpretability, making it a valuable tool for clinical use and early intervention.