Joint Learning of Unsupervised Multi-view Feature and Instance Co-selection with Cross-view Imputation

2025-12-17

AI Summary

Existing methods for unlabeled, incomplete multi-view data treat missing-value imputation, feature selection, and instance selection as disjoint two-stage processes, neglecting their intrinsic interdependence; moreover, naive view fusion fails to capture complementary information across views. Method: We propose the first joint optimization framework that unifies multi-view feature selection, instance selection, and neighborhood-guided cross-view imputation. It leverages sample similarities discovered via co-selection to guide accurate imputation, while improved imputation quality in turn enhances the discriminability of co-selection. The framework incorporates a self-reconstruction objective, cross-view k-nearest-neighbor relationship modeling, multi-view consistency regularization, and an alternating optimization algorithm. Results: Extensive experiments on multiple benchmark datasets demonstrate significant improvements over state-of-the-art methods: both feature subset discriminability and instance representativeness are enhanced simultaneously, and the average imputation error is reduced by 18.7%.

Technology Category

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

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Feature and instance co-selection, which aims to reduce both feature dimensionality and sample size by identifying the most informative features and instances, has attracted considerable attention in recent years. However, when dealing with unlabeled incomplete multi-view data, where some samples are missing in certain views, existing methods typically first impute the missing data and then concatenate all views into a single dataset for subsequent co-selection. Such a strategy treats co-selection and missing data imputation as two independent processes, overlooking potential interactions between them. The inter-sample relationships gleaned from co-selection can aid imputation, which in turn enhances co-selection performance. Additionally, simply merging multi-view data fails to capture the complementary information among views, ultimately limiting co-selection effectiveness. To address these issues, we propose a novel co-selection method, termed Joint learning of Unsupervised multI-view feature and instance Co-selection with cross-viEw imputation (JUICE). JUICE first reconstructs incomplete multi-view data using available observations, bringing missing data recovery and feature and instance co-selection together in a unified framework. Then, JUICE leverages cross-view neighborhood information to learn inter-sample relationships and further refine the imputation of missing values during reconstruction. This enables the selection of more representative features and instances. Extensive experiments demonstrate that JUICE outperforms state-of-the-art methods.

Problem

• Addresses feature and instance co-selection in incomplete multi-view data
• Integrates missing data imputation with co-selection in unified framework
• Captures cross-view complementary information to enhance selection effectiveness

Innovation

• Joint learning of unsupervised multi-view feature and instance co-selection
• Cross-view imputation integrated into a unified framework
• Leverages cross-view neighborhood information for refined imputation