MAMAH, IKECHUKWU COLLINS (2025) DEVELOPMENT OF A PREDICTIVE SYSTEM FOR EARLY DETECTION OF TUBERCULOSIS USING RANDOM FOREST. Other thesis, GODFREY OKOYE UNIVERSITY, ENUGU.
![[img]](https://eprints.gouni.edu.ng/style/images/fileicons/text.png) |
Text
REPO196.docx Download (1MB) |
Abstract
Machine learning technique random forest was explored for predicting tuberculosis. Data from Tuberculosis Dataset for Intelligent and Adaptive Medical Diagnostic System were used to build the prediction model. The main objective is to develop a predictive system for early detection of tuberculosis using random forest which is crucial for early detection before it becomes terminal. A Random forest (RF) model was designed, evaluated and used. According to Ohwo, S (2023) the dataset was gotten through validated and structured questionnaire using random sampling after obtaining the patients' consent. The collated dataset was pre-processed and encoded with variables (features) for prediction which include, loss of energy, night sweat, breathing difficulty, fever, , sputum, immune suppression,, chill, lack of concentration, irritation, chest pain loss of appetite, lymph node enlargement, loss of pleasure systolic blood pressure cough and BMI. Early prediction of tuberculosis from on the clinical data from patients' features using random forest plays a crucial role in the diagnosis, intervention and management of tuberculosis patient. The developed model was evaluated using two performance metrics, including accuracy of 93% and a weighted average of 94%. This results show that the random forest model can confidently be used to predict tuberculosis, promising an alternative to existing systems.
| Item Type: | Thesis (Other) |
|---|---|
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science |
| Divisions: | Faculty of Natural and Applied Sciences |
| Depositing User: | Uchenna Eneogwe |
| Date Deposited: | 19 Jun 2026 13:35 |
| Last Modified: | 19 Jun 2026 13:35 |
| URI: | http://eprints.gouni.edu.ng/id/eprint/5839 |
Actions (login required)
 |
View Item |
