A Survey on Active Feature Acquisition Strategies

2025-02-16

AI Summary

This work addresses prediction tasks in high-cost or high-latency data acquisition scenarios, where the objective is to selectively acquire the most informative features per sample to balance acquisition cost and model accuracy. We systematically survey active feature acquisition (AFA) methods—including uncertainty-based, mutual-information-driven, reinforcement learning–based, and meta-learning–based strategies—and integrate cost-sensitive learning with Bayesian optimization. Our contribution is threefold: (1) we establish the first structured knowledge framework for AFA, unifying problem formulation, evaluation protocols, and challenge taxonomies; (2) we identify critical bottlenecks, notably weak cross-domain generalization and difficulty in modeling dynamic acquisition costs; and (3) we delineate key open problems and future research directions. This framework provides both theoretical foundations and practical guidance for developing efficient, low-cost intelligent decision-making systems.

Technology Category

• Machine Learning: Active Learning
• Constraint Satisfaction and Optimization: Constraint Learning and Acquisition
• Knowledge Representation and Reasoning: Knowledge Acquisition

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics

Abstract

Active feature acquisition studies the challenge of making accurate predictions while limiting the cost of collecting complete data. By selectively acquiring only the most informative features for each instance, these strategies enable efficient decision-making in scenarios where data collection is expensive or time-consuming. This survey reviews recent progress in active feature acquisition, discussing common problem formulations, practical challenges, and key insights. We also highlight open issues and promising directions for future research.

Problem

• Active feature acquisition strategies
• Cost-efficient data collection
• Informative feature selection

Innovation

• Selective feature acquisition
• Cost-efficient data collection
• Informative feature prioritization