Predictive Overlapping Co-Clustering
This work addresses the need for more accurate predictive analysis in applications like healthcare and recommendation systems, though it appears incremental as it builds on existing co-clustering techniques.
The paper tackles the problem of building predictive models with high-dimensional and heterogeneous data by introducing Predictive Overlapping Co-Clustering (POCC), which optimizes co-clusters to maximize predictive power, and results show it outperforms K-means and Spectral co-clustering on four real datasets in terms of precision, recall, and f-measure.
In the past few years co-clustering has emerged as an important data mining tool for two way data analysis. Co-clustering is more advantageous over traditional one dimensional clustering in many ways such as, ability to find highly correlated sub-groups of rows and columns. However, one of the overlooked benefits of co-clustering is that, it can be used to extract meaningful knowledge for various other knowledge extraction purposes. For example, building predictive models with high dimensional data and heterogeneous population is a non-trivial task. Co-clusters extracted from such data, which shows similar pattern in both the dimension, can be used for a more accurate predictive model building. Several applications such as finding patient-disease cohorts in health care analysis, finding user-genre groups in recommendation systems and community detection problems can benefit from co-clustering technique that utilizes the predictive power of the data to generate co-clusters for improved data analysis. In this paper, we present the novel idea of Predictive Overlapping Co-Clustering (POCC) as an optimization problem for a more effective and improved predictive analysis. Our algorithm generates optimal co-clusters by maximizing predictive power of the co-clusters subject to the constraints on the number of row and column clusters. In this paper precision, recall and f-measure have been used as evaluation measures of the resulting co-clusters. Results of our algorithm has been compared with two other well-known techniques - K-means and Spectral co-clustering, over four real data set namely, Leukemia, Internet-Ads, Ovarian cancer and MovieLens data set. The results demonstrate the effectiveness and utility of our algorithm POCC in practice.