A meta-heuristic density-based subspace clustering algorithm for high-dimensional data

Subspace clustering is one of the efficient techniques for determining clusters in different subsets dimensions. Ideally, these should find all possible non-redundant which data point participates. Unfortunately, existing hard subspace algorithms fail to satisfy this property. Additionally, with increase dimensions data, classical become inefficient. This work presents a new density-based algorithm (S_FAD) overcome drawbacks algorithms. S_FAD based on bottom-up approach and finds varied density using parameters DBSCAN algorithm. The optimizes DBCAN through hybrid meta-heuristic uses hashing concepts discover clusters. efficacy evaluated against various artificial real datasets terms F_Score rand_index. Performance assessed three parameters: average ranking, SRR scalability Statistical analysis performed Wilcoxon signed-rank test. Results reveal that performs considerably better majority scales well up 6400 actual dataset.