Tri-level Robust Clustering Ensemble with Multiple Graph Learning

Clustering ensemble generates a consensus clustering result by integrating multiple weak base results. Although it often provides more robust results compared with single methods, still suffers from the robustness problem if does not treat unreliability of carefully. Conventional methods use all data for ensemble, while ignoring noises or outliers on data. some are proposed, which extract data, they characterize in level, and thus cannot comprehensively handle complicated problem. In this paper, to address problem, we propose novel Tri-level Robust Ensemble (TRCE) method transforming graph learning Just as its name implies, proposed tackles three levels: level instance level. By considering comprehensive way, TRCE can achieve result. Experimental benchmark datasets also demonstrate it. Our outperforms other state-of-the-art methods. Even ours performs better.