Data aggregation from different database into a data warehouse creates multi-dimensional data such as data cubes. With regard the 3D structure of data, data cube clustering has significant challenges to perform on data cube. In this paper, new preprocessing techniques and a novel hybridization of DBSCAN and fuzzy Earthworm Optimization Algorithm (EWOA) are proposed to solve the challenges. Proposed preprocessing consists of an assigned address to each cube cell, dimension move to create a related 2D data from the data cube and new similarity metric. The DBSCAN algorithm, as a density-based clustering algorithm, is adopted based on both Euclidean and new proposed similarity metric, which call DBSCAN1 and DBSCAN2 for the related 2D data. A new hybridization of the EWOA and DBSCAN is proposed to improve the DBSCAN and called EWOA-DBSCAN. Also, to dynamically tune parameters of EWOA, a fuzzy logic controller (FLC) is designed with two fuzzy groups rules of Mamdani (EWOA-DBSCAN-Mamdani) and Sugeno (EWOA-DBSCAN-Sugeno), separately. These ideas are proposed to present efficient and flexible unsupervised analysis for a data cube by utilizing a meta-heuristic algorithm to optimize DBSCAN’s parameters and increasing the efficiency of the idea by applying dynamic tuning parameters of the algorithm. To evaluate the efficiency, the proposed algorithms are compared with DBSCAN1 and GA-DBSCAN1, GA-DBSCAN1-Mamdani and GA-DBSCAN1-Sugeno. The experimental results, consisted of 20 times running, indicate that the proposed ideas achieved to their targets.
پیوند مجله / همایش