An enhanced human learning optimisation algorithm for effective data clustering

Clustering arranges the data objects into clusters and similar data objects put into same cluster. There is not a single algorithm that can work effectively with all types of clustering problems. Other side, quality of clusters is also an important issue for clustering algorithms. This work deals with aforementioned issues and presents a new clustering algorithm-based human learning optimisation. Several modifications are incorporated into HLO algorithm for alleviating similar ability of individual learning and control parameter issues, called enhanced HLO (EHLO). The similar learning ability of individuals can be enhanced based on learner phase of TLBO algorithm. The control parameter issue of random learning is resolved through logistic chaotic map. The well-known clustering data sets are chosen for conducting the experiments and results are compared to eight state-of-the-art clustering algorithms based on several performance metrics. The results showed that EHLO obtains superior clustering results than other algorithms. The performance is also assessed through Friedman statistical test and it is followed by a post-hoc test. Results indicated that the proposed EHLO algorithm obtains an average rank of 1.29 compared to other algorithms, the p-value (3.71E-04) and the critical value (16.919). Hence, the proposed EHLO is validated through experimental and statistical results.