Guided Manifold Alignment with Geometry-Regularized Twin Autoencoders
This work addresses a real-world limitation in manifold alignment for multi-modal data integration, with incremental improvements in generalization and robustness.
The paper tackles the problem of manifold alignment's inability to generalize to unseen data by proposing a guided framework with geometry-regularized twin autoencoders, showing improvements in embedding consistency and cross-domain transfer, and applying it to Alzheimer's disease diagnosis to enhance predictive accuracy.
Manifold alignment (MA) involves a set of techniques for learning shared representations across domains, yet many traditional MA methods are incapable of performing out-of-sample extension, limiting their real-world applicability. We propose a guided representation learning framework leveraging a geometry-regularized twin autoencoder (AE) architecture to enhance MA while enabling generalization to unseen data. Our method enforces structured cross-modal mappings to maintain geometric fidelity in learned embeddings. By incorporating a pre-trained alignment model and a multitask learning formulation, we improve cross-domain generalization and representation robustness while maintaining alignment fidelity. We evaluate our approach using several MA methods, showing improvements in embedding consistency, information preservation, and cross-domain transfer. Additionally, we apply our framework to Alzheimer's disease diagnosis, demonstrating its ability to integrate multi-modal patient data and enhance predictive accuracy in cases limited to a single domain by leveraging insights from the multi-modal problem.