Deep Approximately Orthogonal Nonnegative Matrix Factorization for Clustering
This work addresses clustering tasks in domains like image analysis, but it is incremental as it builds on existing deep NMF methods by adding orthogonality constraints.
The paper tackled the problem of improving clustering performance by proposing a deep approximately orthogonal nonnegative matrix factorization method that imposes both nonnegativity and orthogonality for hierarchical clustering. The result showed better clustering performance on two face image datasets compared to other deep matrix factorization methods and state-of-the-art single-layer NMF variants.
Nonnegative Matrix Factorization (NMF) is a widely used technique for data representation. Inspired by the expressive power of deep learning, several NMF variants equipped with deep architectures have been proposed. However, these methods mostly use the only nonnegativity while ignoring task-specific features of data. In this paper, we propose a novel deep approximately orthogonal nonnegative matrix factorization method where both nonnegativity and orthogonality are imposed with the aim to perform a hierarchical clustering by using different level of abstractions of data. Experiment on two face image datasets showed that the proposed method achieved better clustering performance than other deep matrix factorization methods and state-of-the-art single layer NMF variants.