Explore the Design and Development of an intelligent multi-output Hybrid Advisory Model for Medical Diagnosis

Authors

  • Abraham Alamirew Debark University Author
  • Birhanu Alemneh Tiruwork Debark University Author
  • Yeshiwas Adane Debark University Author

Abstract

Healthcare medical diagnosis is the process of identifying a disease, condition, or injury based on a patient's symptoms, physical examination, and diagnostic tests. The AI-based advisory model utilizes artificial intelligence techniques to provide recommendations, guidance, and decision-making support across various domains. This study explores the development and evaluation of a multi-output hybrid advisory model for healthcare medical diagnosis. Multi-Output Classifier models analyze datasets to aid healthcare decisions, helping diagnose conditions, suggest treatments, and provide medical information. We used a kaggle dataset (16,125) and an additional dataset (32,631) that was meticulously curated from 12 trusted websites of the National Institutes of Health (NIH), an official trusted website of the United States government, to build a medical diagnosis model via six experiments. The dataset was partitioned into 80% for training and 20% for testing. To optimize performance, we have developed a hybrid model that combines the strengths of individual models to create a more robust model that overcomes their individual limitations. We evaluate the model classification performance using metrics such as accuracy, precision, recall, F1 score, and cross-validation. Additionally, we evaluated the model's performance with BERT and confidence scores for question answering. Six experiments were conducted using Extra Trees, random forests, LSTM, MLP, a hybrid model, and Extreme Gradient Boosting. The hybrid model achieved the highest accuracy of 97%. As a result, the hybrid model was chosen as the best for medical diagnosis. The proposed multi-output hybrid model achieved an acceptance rate of 84% among domain experts.

Author Biographies

  • Birhanu Alemneh Tiruwork, Debark University

    Computer Science,Lecturer

  • Yeshiwas Adane, Debark University

    Computer Science,Lecturer

Downloads

Published

2026-11-13

Issue

Section

Articles

How to Cite

Explore the Design and Development of an intelligent multi-output Hybrid Advisory Model for Medical Diagnosis. (2026). Ras Dejen Journal of Sciences, 1(2). https://www.dkujournals.edu.et/index.php/RDJS/article/view/66