Rough Set Based Similarity Measures for Data Analytics in Spatial Epidemiology

Rough Set Based Similarity Measures for Data Analytics in Spatial Epidemiology

Sharmila Banu K., B.K. Tripathy
Copyright: © 2016 |Pages: 10
DOI: 10.4018/IJRSDA.2016010107
(Individual Articles)
No Current Special Offers


Epidemiological studies are carried out to understand the pattern and transmission of disease instances. Some prominent dimensions considered for analysis are cohort studies, ecological studies, transmission modeling and prediction. ‘Descriptive Epidemiology' is defined with respect to ‘people, time and place'. Place (geography) plays a key role in the pattern of disease outcomes in both epidemic outbreaks and chronic cases. A lot of research has documented the significance of spatial features in Epidemiology and have produced health/disease maps of a particular geography. This work proposes to identify similarity between regions in such maps using Rough set based measures. Thus spatial auto-correlation of disease instances in a geographic region can be analysed further to prepare mitigation strategies.
Article Preview

1. Introduction

Mankind has been battling diseases from ancient to modern times across all civilizations. Wandering from place to place, settling on the river banks, adapting to industrialization till recent lifestyle has led us through a myriad category of diseases and reasons. Epidemic outbreaks have been studied in reference to the geographic areas where they originate and spread. It began with the study of cholera instances in London during early 1800s when John Snow identified the source of infection to a common water pump using a map. Hippocrates, Fracastoro, Sir Edwin Chadwick and others have laid out the doctrines on which modern epidemiology is based. These forerunners used statistics, demographics and maps to understand interrelationships between humans, diseases, geographies and other root causes. Modern Epidemiology deals with more detailed analysis of these factors using sophisticated techniques with the objective of identifying potential interventions to save countless lives. Extreme climatic conditions, exposure to natural and man-made factors that are not conducive to good health and prevalence social imbalances pose serious challenges to public health. With the use of GIS (Geographic Information System) where data are tagged geographically, studying factors involved in disease complications has become sophisticated. Developing countries need to identify its potential as well as weak links in terms of distribution of diseases. A public domain with information on health census of the population is essential. This work will involve dealing with uncertainties in data across regions.

A lot of computer science based mining algorithms have been developed to efficiently mine medical data and a new branch of science called Medical Informatics has evolved. Infrastructures for Health-care services are getting established and plans for efficient deliveries are being worked upon especially in developed countries. Recent research works call for information technology in health care services. They discuss the application of information technology for health infrastructure. Healthcare domain is fortified with advanced Image processing and analyses like digital watermarking of intravascular ultrasound video documented in N.Dey et al. (2012), firefly algorithm for ophthalmology imaging in N.Dey et al. (2014).

Soft computing techniques involving Fuzzy Sets, Rough Sets, Neighborhood Rough Sets and their variants are used to analyse medical and epidemiological data. These concepts are extensions of classical set theory. They provide the sophistication of dealing with vague, incomplete and uncertain data to come up with useful patterns. Rough Sets use approximate reasoning to uncover useful information from data. Rough set theory has been used for Medical diagnosis and outcome prediction. Rough set based analytics of epidemiological data has been carried out which involves medical, demographical and spatial data.

The remaining part of the paper is organized with section 2 elaborating on the spatial epidemiology and discusses the relevant literature, section 3 discusses the spatial correlation in spatial data analysis and how can this be addressed using the similarity measure in rough sets, section 4 outlines the preliminaries of rough set theory and elaborates on RST based similarity measures to find similar affected regions in a given geography followed by results and discussions and the last section presents the concluding remarks of the paper.

Complete Article List

Search this Journal:
Volume 9: 1 Issue (2025): Forthcoming, Available for Pre-Order
Volume 8: 1 Issue (2024): Forthcoming, Available for Pre-Order
Volume 7: 4 Issues (2021): 1 Released, 3 Forthcoming
Volume 6: 3 Issues (2019)
Volume 5: 4 Issues (2018)
Volume 4: 4 Issues (2017)
Volume 3: 4 Issues (2016)
Volume 2: 2 Issues (2015)
Volume 1: 2 Issues (2014)
View Complete Journal Contents Listing