Authors:
Daranee Thitiprayoonwongse
1
;
Nuanwan Soonthornphisaj
1
and
Prapat Suriyaphol
2
Affiliations:
1
Faculty of Science Kasetsart University, Thailand
;
2
Siriraj Hospital Mahidol University, Thailand
Keyword(s):
Data mining, Decision tree, Dengue virus disease.
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence
;
Data Mining
;
Databases and Information Systems Integration
;
Enterprise Information Systems
;
Sensor Networks
;
Signal Processing
;
Soft Computing
Abstract:
Dengue infection is an epidemic disease typically found in tropical region. Symptoms of the disease show rapid and violent for patients in a short time. The World Health Organization (WHO) classifies the dengue infection as Dengue Fever (DF) and Dengue Hemorrhagic Fever (DHF). Symptoms of DHF are divided into 4 types. The problem might be happen when an expert misdiagnoses dengue infection. For Example, an expert diagnosed a patient as non dengue or DF even if a patient was a DHF patient. That might be the cause of dead if patient did not receive treatment. Therefore, we selected data mining approach to solve this problem. We employed decision tree algorithm to learn from data set in order to create new knowledge. The first experimental result shows useful knowledge to classify dengue infection levels into 4 groups (DF, DHF I, DHF II, and DHF III). An average accuracy is 96.50 %. The second experimental result shows the tree and a set of rules to classify dengue infection levels int
o 2 groups followed by our assumption. An accuracy is 96.00 %. Furthermore, we compared our performance in term of false negative values to WHO and some researchers and found that our research outperforms those criteria, as well.
(More)