Ant Colony Based Optimization from Infrequent Itemsets

摘要

Data Mining is the area of research by which we can find relevant patterns from the data set. It is used in several areas. In this paper we are focusing on finding relevant patterns from Positive and Negative Rules. For this we have applied Ant Colony Optimization (ACO) technique on the positive and negative rules. Our algorithm has achieved better global optimum value and chances of finding are improved. So the chances of Positive or relevant rules are more in comparison to the traditional technique. We are also applying the optimization to the negative rules so that there are equal chances for achieving the global optimum. But the negative rules are not qualifying the global optimum value and hence the relevant rules find by our algorithm are verified.