文章基本信息

标题：Fuzzy logic to analyze survey data from populations exposed to arsenic-contaminated water
本地全文：下载
作者：Jose Arturo Molina Mora
期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2016
卷号：5
期号：1
页码：0095-0098
出版社：Shri Pannalal Research Institute of Technolgy
摘要：Drinking water contaminated with arsenic is a public health problem that currently affects about 11,000 people in Costa Rica, mainly in the north and northwest of the country which source of water is wells-based. Poisoning of this water has, both chronic and acute, major impact on population health and lead to clinical manifestations of neurological, dermal, gastrointestinal and hematological level, among others. Thus, water for subsistence taken from wells of areas with arsenic contamination must be constantly monitored. In this work, a set of data from surveys, including factors such as the degree of contamination of wells that currently use, education, membership of a community association and the distance of a new well, were used for analyzing the intent of change of well. For this, a model of fuzzy logic was applied to handle uncertainty and create rules of association between variables. Also, a comparison was made with classic algorithms including K-means algorithms, decision trees and neural networks. The resolution of the models did not differ from each other and all with about 70% accuracy. To improve the response capacity, a factor analysis was performed using principal component analysis (PCA), obtaining an accuracy above 98% in all cases, including fuzzy logic. Thus, the models using fuzzy logic tools are presented as alternatives to data processing in data mining.
关键词：Fuzzy-logic; Drinking Water; Data Mining