Water quality monitoring based on support vector machine at Putrajaya lake and wetland / Zakirah Binti Mohd Azamli
A support vector machine (SVM) model is a machine learning technology. It is used for the proposed to classify the data by recognize the patterns of data. It is by work by train the data of sampling area to predict the water quality of Putrajaya Lake and Wetland based from data of water quality para...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2012
|
| Subjects: | |
| Online Access: | http://studentsrepo.um.edu.my/3717/4/1._Title_page%2C_abstract%2C_content.pdf http://studentsrepo.um.edu.my/3717/5/2._Chap_1_%E2%80%93_6.pdf http://studentsrepo.um.edu.my/3717/6/3._References.pdf http://studentsrepo.um.edu.my/3717/7/4._Appendices.pdf http://pendeta.um.edu.my/client/default/search/results?qu=Water+quality+monitoring+based+on+support+vector+machine+at+Putrajaya+lake+and+wetland&te= http://studentsrepo.um.edu.my/3717/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | A support vector machine (SVM) model is a machine learning technology. It is used for the proposed to classify the data by recognize the patterns of data. It is by work by train the data of sampling area to predict the water quality of Putrajaya Lake and Wetland based from data of water quality parameters of 29 sampling stations in study area. The following seven water quality parameters were used for the proposed of SVM classification, namely; dissolved oxygen (D.O), temperature (Temp), water pH, salinity, Biochemical Oxygen Demand dan Escherichia coli (E. coli). The dissolved oxygen variable is being used as indicator of Putrajaya Lake and Wetland water quality measurements standard. The propose
of this study is to predict the dissolve oxygen based on three level of class of water quality namely; High, Medium and Low. The data were undergoing training and testing. RBF
kernel function is employed to train the data. The result shows that cross validation error is 0.304762. The optimal cost C and sigma value are 0.7 and 0.5. Meanwhile, the numbers of support vector on the other hand are about 90. The resulting from sensitivity and specificity is also determined. |
|---|