Cross-Linked Unified Embedding for cross-modality representation learning

Multi-modal learning is essential for understanding information in the real world. Jointly learning from multi-modal data enables global integration of both shared and modality-specific information, but current strategies often fail when observa- tions from certain modalities are incomplete or missing for part of the subjects. To learn comprehensive representations based on such modality-incomplete data, we present a semi-supervised neural network model called CLUE (Cross-Linked Unified Embedding). Extending from multi-modal VAEs, CLUE introduces the use of cross-encoders to construct latent representations from modality-incomplete observations. Representation learning for modality-incomplete observations is common in genomics.