Use of SVM Technique for Comparison of Water Quality: A Case Study from China
Volume 11, Number 5, September 2015 - Paper 7 - pp. 473-480
JINMEI LU1, FUQING YUAN1, FENGYING ZHANG21 Department of Engineering and Safety, University of Tromsø, N-9037 Tromsø, NORWAY
2 China national Environmental Monitoring Centre, Beijing 100012, CHINA
(Received on December 10, 2014, revised on June 15, 2015)
Trace elements in drinking water can influence the longevity of human beings. Drinking water samples from a longevity county Xiayi and a nearby non-longevity county Yucheng were collected and tested the element concentrations. Support vector machine (SVM), as a pattern recognition technique for small sample size classification problem, was applied to analyze the difference of element concentrations in the drinking water from both counties. Further the SVM is used to select the elements with the most discrepancy. The study revealed large discrepancies in element concentrations in drinking water from these two counties. Barium, Ca, F, Mg, Mn, Si and Sr in drinking water from Xiayi is higher than that from Yucheng, whereas the Fe, K, Pb, Cu, Co, Cr, Ni in drinking water from Xiayi is lower. The significance of difference is Cr> Mn> Co> Fe. The high Mn, Ca and low Cr, Cd, Co, Fe contents in drinking water from Xiayi are conductive to local longevity.
Click here to download the paper.
Please note : You will need Adobe Acrobat viewer to view the full articles.