Optimised Clustering Association Rule Mining with Health Data

Clustered Particle Swarm Optimisation (PSO) based approaches can be used to extract strong and higher-order association rules. This paper illustrates the application of clustering to PSO-based association rule mining. This work investigates improvements to association rules and introduces an algorithm to perform PSO-Based association rule mining integrated with a clustering approach. The proposed algorithm does not require any user-defined threshold or parameters to extract the strongest rules. To measure the performance of the proposed algorithm, we used data on lung cancer and disease symptoms. The proposed algorithm shows promising results in terms of generating strong rules and reducing execution time from the experimental results and the comparison study. This research can be used to investigate the strong associations among diseases and related attributes. However, the application of the proposed method is not limited to the health and medical domain, and it can also be employed in other domains to extract associations among class levels and other attributes.