Regularizing Autoencoder-Based Matrix Completion Models via Manifold Learning
This work addresses a specific problem in matrix completion for recommendation systems or data imputation, but it is incremental as it builds on existing autoencoder models with a regularization approach.
The paper tackles overfitting in autoencoder-based matrix completion when training data is scarce by introducing a data-dependent regularization technique using manifold learning as an auxiliary task, resulting in improved reconstruction accuracy on well-known datasets.
Autoencoders are popular among neural-network-based matrix completion models due to their ability to retrieve potential latent factors from the partially observed matrices. Nevertheless, when training data is scarce their performance is significantly degraded due to overfitting. In this paper, we mit- igate overfitting with a data-dependent regularization technique that relies on the principles of multi-task learning. Specifically, we propose an autoencoder-based matrix completion model that performs prediction of the unknown matrix values as a main task, and manifold learning as an auxiliary task. The latter acts as an inductive bias, leading to solutions that generalize better. The proposed model outperforms the existing autoencoder-based models designed for matrix completion, achieving high reconstruction accuracy in well-known datasets.