A Modified Redescriptive Data Mining Technique to Relate Cyclones and Temperatures of Coastal AP, India

Redescription Mining, not a new problem describes set of objects or entities in at least two ways either with Boolean, Categorical or Real data. With the consideration of real valued data, the constraints that are applied with some range of values given more scope to the accuracy of redescriptions and that are to be statistically significant. Considering Geographical area, pertained to some specific regions of Andhra Pradesh, India, in this paper we consider the Natural Hazards and the effect of Temperature on those areas that can describe the regions in two ways, either by the variations in temperature or by the effect and severity of natural hazards. An attempt is made in this paper, to show the correlation between frequency of Natural hazards and temperature increase in the coastal areas of Andhra Pradesh. The algorithm proposed works for real-data on either side of the description. Nine coastal districts of Andhra Pradesh are considered. Study on these districts shows how these districts are more vulnerable to Cyclones in specific, when compared to other districts. Experiments on this resulted in the redescription of coastal areas with temperature and cyclones.