Generation of temporal class association rules from quantitative data using evolutionary approach

Most of the data mining algorithms perform analysis on quantitative data only after performing discretization. Nowadays, there is a great interest in finding the health impacts of climate change. One of the factors that cause changes in the climate is the ozone layer. Adverse levels of ozone may cause several diseases like asthma, chronic disorders and other respiratory symptoms. Hereby we present an evolutionary approach based association technique to find the relationship between several multidimensional climatological variables that are involved in determining an ozone day. The relationships between variables are discovered by generating quantitative association rules that exhibit a temporal pattern. When association rules are generated from high dimensional quantitative databases, the rules suffer from loss of information due to discretization. To overcome this problem, the proposed approach involves genetic algorithm to discover all possible dependencies between variables with optimal intervals. Our method generates quantitative association rules on temporal database, with more realistic interval rather than crisp boundary.