Doubly Non-Central Beta Matrix Factorization for Stable Dimensionality Reduction of Bounded Support Matrix Data
This work addresses the need for stable dimensionality reduction in domains like bioinformatics, where bounded support data is common, though it is incremental as it builds on existing Tucker decomposition methods.
The paper tackled the problem of interpretable and computationally efficient matrix decomposition for bounded support data, such as in DNA methylation studies, by developing a Tucker decomposition method with a sampling algorithm. The result showed similar prediction and computational performance to state-of-the-art approaches but significantly better stability to hyper-parameter changes, leading to higher confidence in applications like cancer analysis.
We consider the problem of developing interpretable and computationally efficient matrix decomposition methods for matrices whose entries have bounded support. Such matrices are found in large-scale DNA methylation studies and many other settings. Our approach decomposes the data matrix into a Tucker representation wherein the number of columns in the constituent factor matrices is not constrained. We derive a computationally efficient sampling algorithm to solve for the Tucker decomposition. We evaluate the performance of our method using three criteria: predictability, computability, and stability. Empirical results show that our method has similar performance as other state-of-the-art approaches in terms of held-out prediction and computational complexity, but has significantly better performance in terms of stability to changes in hyper-parameters. The improved stability results in higher confidence in the results in applications where the constituent factors are used to generate and test scientific hypotheses such as DNA methylation analysis of cancer samples.