Home / Collection / College of Computing and Informatics / Information Technology / View Item
This study aims to identify human diseases in Amharic based on observed symptoms. The research seeks to enhance the accuracy of disease identification and empower communities to independently monitor their health conditions. It addresses significant health challenges in Ethiopia, where a notable shortage and misdistribution of healthcare workers, particularly doctors, impede effective patient treatment. With both communicable and non-communicable diseases contributing substantially to mortality rates, there is an urgent need for innovative diagnostic solutions, especially in rural areas with limited access to healthcare facilities. This study proposes the development of utilizing machine learning for symptom-based identification of human disease specifically designed for the Amharic language. Eight machine learning algorithms were utilized in this study. Among them, the Gradient Boosting algorithm achieved the highest accuracy of 87.23% using term frequency-inverse document frequency (TF-IDF) for feature extraction. The findings demonstrate the feasibility of applying machine learning algorithms for disease identification in the Amharic language. Furthermore, the study recommends expanding the range of diseases identified and exploring the potential of deep learning algorithms for future research. KEYWORD: Machine
 |
Name: 3fcf719c0f6d21bbe984c0ef9b97b414.pdf | Download |
| Size: 2.23 Mb | ||
| Format: pdf |