Semi-Supervised Regression using Cluster Ensemble and Low-Rank Co-Association Matrix Decomposition under Uncertainties
This addresses semi-supervised regression for uncertain data, but appears incremental as it combines existing regularization and ensemble techniques with computational optimizations.
The paper tackles semi-supervised regression under data uncertainty by combining graph Laplacian regularization with cluster ensemble methods, using a low-rank decomposed co-association matrix as a similarity measure. Numerical experiments demonstrate improved robustness, efficiency, and scalability.
In this paper, we solve a semi-supervised regression problem. Due to the lack of knowledge about the data structure and the presence of random noise, the considered data model is uncertain. We propose a method which combines graph Laplacian regularization and cluster ensemble methodologies. The co-association matrix of the ensemble is calculated on both labeled and unlabeled data; this matrix is used as a similarity matrix in the regularization framework to derive the predicted outputs. We use the low-rank decomposition of the co-association matrix to significantly speedup calculations and reduce memory. Numerical experiments using the Monte Carlo approach demonstrate robustness, efficiency, and scalability of the proposed method.