A Case Based Reasoning Knowledge Based System for Hypertension Management
| dc.contributor.advisor | Kebede, Gashaw (PhD) | |
| dc.contributor.author | Bekele, Henok | |
| dc.date.accessioned | 2018-11-27T14:15:48Z | |
| dc.date.accessioned | 2023-11-29T04:56:53Z | |
| dc.date.available | 2018-11-27T14:15:48Z | |
| dc.date.available | 2023-11-29T04:56:53Z | |
| dc.date.issued | 2011-07 | |
| dc.description.abstract | Hypertension is a growing public health problem. In this paper, the potential of case based reasoning approach for hypertension management have been investigated. In order to conduct the research, the required knowledge for the study have been collected from hypertension compliant card histories, domain experts and other relevant documents through semi-structured interview and document analysis methods of knowledge elicitation. Then the knowledge is modeled in hierarchical tree manner and case structure of the case base is constructed. Forty five hypertension cases are collected from Brook Medical Service Plc and Bole 17 Health Center to construct the case base. The case based reasoning prototype for hypertension management is implemented by using python 2.6. Nearest Neighbor retrieval algorithm, voting method, domain expert feedback and incremental learning are used for the retrieval, reuse, revise and retain tasks of the prototype respectively. The collected hypertension cases are represented in the form of Feature-vector case representation approaches. The prototype is evaluated by using both statistical analysis and user evaluation. The statistical analysis uses leave-one-out cross validation testing proportion for both the retrieval and reuse processes. The retrieval performance of the prototype shows average value of 86.1% recall and 60% precision, while the performance of the reuse process shows an average value of 88.89% accuracy. The over all performance of the prototype as it is evaluated by domain experts is 3.99 out of 5. Given all these results, the performance of the prototype is promising. All in all the study achieves its objective by developing the prototype with promising performance and user acceptance, and demonstrating case based reasoning approach in designing knowledge based system for hypertension management. | en_US |
| dc.identifier.uri | http://etd.aau.edu.et/handle/123456789/14573 | |
| dc.language.iso | en | en_US |
| dc.publisher | Addis Ababa University | en_US |
| dc.subject | Knowledge Based System | en_US |
| dc.subject | for Hypertension Management | en_US |
| dc.title | A Case Based Reasoning Knowledge Based System for Hypertension Management | en_US |
| dc.type | Thesis | en_US |