![[2602.16709] Knowledge-Embedded Latent Projection for Robust Representation Learning](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.16709] Knowledge-Embedded Latent Projection for Robust Representation Learning
Summary
This article presents a novel knowledge-embedded latent projection model aimed at improving representation learning in high-dimensional data, particularly in electronic health records (EHRs).
Why It Matters
The research addresses the challenges of representation learning in imbalanced datasets, which is crucial for fields like healthcare where data availability can be limited. By leveraging external semantic embeddings, the proposed model enhances the robustness and accuracy of data analysis, potentially improving patient outcomes and research insights.
Key Takeaways
- Introduces a knowledge-embedded latent projection model for robust representation learning.
- Addresses challenges in high-dimensional data analysis, particularly in EHRs.
- Utilizes external semantic embeddings to enhance model performance.
- Develops a computationally efficient estimation procedure combining kernel PCA and gradient descent.
- Provides theoretical guarantees on estimation error and convergence for the proposed method.
Computer Science > Machine Learning arXiv:2602.16709 (cs) [Submitted on 18 Feb 2026] Title:Knowledge-Embedded Latent Projection for Robust Representation Learning Authors:Weijing Tang, Ming Yuan, Zongqi Xia, Tianxi Cai View a PDF of the paper titled Knowledge-Embedded Latent Projection for Robust Representation Learning, by Weijing Tang and Ming Yuan and Zongqi Xia and Tianxi Cai View PDF HTML (experimental) Abstract:Latent space models are widely used for analyzing high-dimensional discrete data matrices, such as patient-feature matrices in electronic health records (EHRs), by capturing complex dependence structures through low-dimensional embeddings. However, estimation becomes challenging in the imbalanced regime, where one matrix dimension is much larger than the other. In EHR applications, cohort sizes are often limited by disease prevalence or data availability, whereas the feature space remains extremely large due to the breadth of medical coding system. Motivated by the increasing availability of external semantic embeddings, such as pre-trained embeddings of clinical concepts in EHRs, we propose a knowledge-embedded latent projection model that leverages semantic side information to regularize representation learning. Specifically, we model column embeddings as smooth functions of semantic embeddings via a mapping in a reproducing kernel Hilbert space. We develop a computationally efficient two-step estimation procedure that combines semantically guided subspace c...