|
Hierarchical Clustering Analysis for Specific Liver Disease-Dengue Hemorrhagic Fever (DHF)
|
|
| Author:
|
M.KARPAGAM , DR.K. GEETHA, DR.C. RAJAN
|
| Abstract:
|
Dengue fever is a viral infection transmitted to humans by mosquitoes that live in tropical and subtropical climates and carry the virus. Blood testing detects the dengue virus or antibodies produced in response to dengue infection. Nowadays, the doctors need to know the set of predicted features on dengue virus in order to classify the infected patients and suspected patients, where the suspected patients can be treated in advance by predicting the similar features from the infected patients. This works applies data mining techniques on the real-time clinical data. The data consists of the infected patient records from their admitted date to death date. The sources of data are collected from Rajapalayam Hospital which is a crucial region of Tamilnadu where Dengue death rate is raised. Each dataset consists of nearly 200 attributes. To achieve the knowledge discovery task, this work concentrates on Hierarchical clustering as a data mining technique. The proposed work includes 3 processes to validate the Dengue virus. The first process Preprocessing involves filtering the risk factors for developing dengue hemorrhagic fever. The second concentrates on eliminating trivial factors of disease from the infected persons .The third process involves the final stage in validating the virus from the calculated IgG(Immunoglobulin G) and IgM(Immunoglobulin M ) values from the clinical dataset with initial results(attributes regarding blood) from the laboratory test and resulted dendrogram shows the age factor and infected feature . The high IGM is the critical stage of dengue patients that some patients face the fatal condition. Therefore, the experts can utilize the method to predict the virus based on the symptoms intense.
|
| Keyword:
|
Data Mining, Dengue infection, IGm and IGg
|
| EOI:
|
-
|
| DOI:
|
https://doi.org/10.31838/ijpr/2020.SP1.388
|
| Download:
|
Request For Article
|
|
|