Learned Robust PCA: A Scalable Deep Unfolding Approach for High-Dimensional Outlier Detection
This work addresses outlier detection in low-rank matrix reconstruction for machine learning applications, offering an incremental improvement in efficiency and performance over existing RPCA algorithms.
The paper tackles the problem of high-dimensional outlier detection in robust principal component analysis (RPCA) by proposing Learned Robust PCA (LRPCA), a scalable and learnable non-convex approach using deep unfolding, which outperforms state-of-the-art methods like ScaledGD and AltProj in numerical experiments.
Robust principal component analysis (RPCA) is a critical tool in modern machine learning, which detects outliers in the task of low-rank matrix reconstruction. In this paper, we propose a scalable and learnable non-convex approach for high-dimensional RPCA problems, which we call Learned Robust PCA (LRPCA). LRPCA is highly efficient, and its free parameters can be effectively learned to optimize via deep unfolding. Moreover, we extend deep unfolding from finite iterations to infinite iterations via a novel feedforward-recurrent-mixed neural network model. We establish the recovery guarantee of LRPCA under mild assumptions for RPCA. Numerical experiments show that LRPCA outperforms the state-of-the-art RPCA algorithms, such as ScaledGD and AltProj, on both synthetic datasets and real-world applications.