Interestingness Measure for Mining Spatial Gene Expression Data using Association Rule (1001.3498v1)

Abstract: The search for interesting association rules is an important topic in knowledge discovery in spatial gene expression databases. The set of admissible rules for the selected support and confidence thresholds can easily be extracted by algorithms based on support and confidence, such as Apriori. However, they may produce a large number of rules, many of them are uninteresting. The challenge in association rule mining (ARM) essentially becomes one of determining which rules are the most interesting. Association rule interestingness measures are used to help select and rank association rule patterns. Besides support and confidence, there are other interestingness measures, which include generality reliability, peculiarity, novelty, surprisingness, utility, and applicability. In this paper, the application of the interesting measures entropy and variance for association pattern discovery from spatial gene expression data has been studied. In this study the fast mining algorithm has been used which produce candidate itemsets and it spends less time for calculating k-supports of the itemsets with the Boolean matrix pruned, and it scans the database only once and needs less memory space. Experimental results show that using entropy as the measure of interest for the spatial gene expression data has more diverse and interesting rules.