A Semi-supervised Generative Model for Incomplete Multi-view Data Integration with Missing Labels

2025-08-14

AI Summary

To address the dual challenges of view incompleteness and label scarcity in multi-view data, this paper proposes a semi-supervised generative multi-view learning model. Methodologically, it jointly optimizes an information bottleneck objective on labeled data and a variational likelihood objective on unlabeled data to construct a shared latent space; enforces multi-view consistency via cross-view mutual information maximization; and employs a product-of-experts architecture to ensure robustness against arbitrary view-missing patterns. The work innovatively extends the information bottleneck principle to a semi-supervised generative framework, seamlessly integrating variational inference with multi-view consistency constraints. Experiments on image and multi-omics benchmark datasets demonstrate that the model significantly outperforms state-of-the-art methods in both classification accuracy and missing-view imputation quality—particularly under extreme label scarcity.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Computer Vision: Multi-modal Vision
• Constraint Satisfaction and Optimization: Constraint Learning and Acquisition

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs

Abstract

Multi-view learning is widely applied to real-life datasets, such as multiple omics biological data, but it often suffers from both missing views and missing labels. Prior probabilistic approaches addressed the missing view problem by using a product-of-experts scheme to aggregate representations from present views and achieved superior performance over deterministic classifiers, using the information bottleneck (IB) principle. However, the IB framework is inherently fully supervised and cannot leverage unlabeled data. In this work, we propose a semi-supervised generative model that utilizes both labeled and unlabeled samples in a unified framework. Our method maximizes the likelihood of unlabeled samples to learn a latent space shared with the IB on labeled data. We also perform cross-view mutual information maximization in the latent space to enhance the extraction of shared information across views. Compared to existing approaches, our model achieves better predictive and imputation performance on both image and multi-omics data with missing views and limited labeled samples.

Problem

• Integrating incomplete multi-view data with missing labels
• Leveraging unlabeled data in semi-supervised generative framework
• Enhancing cross-view information extraction with mutual information maximization

Innovation

• Semi-supervised generative model with unified framework
• Latent space learning via likelihood maximization
• Cross-view mutual information maximization technique