Enhanced Multi-model Online Conformal Prediction

2026-01-04

arXiv.org

AI Summary

This work addresses the challenge in online conformal prediction where a single model struggles to consistently maintain prediction set efficiency, while existing multi-model approaches incur high computational costs and are susceptible to interference from poorly performing models. To overcome these limitations, we propose a novel multi-model online conformal prediction algorithm that dynamically selects an efficient subset of models at each time step by constructing a bipartite graph and then chooses the optimal model from this subset to generate compact prediction sets. By avoiding exhaustive search over all candidate models, our method significantly reduces computational complexity while preserving the validity of prediction sets. Experimental results demonstrate that the proposed approach outperforms existing methods in both prediction set size and computational efficiency, thereby enhancing the performance of online uncertainty quantification.

Technology Category

• Machine Learning: Online Learning & Bandits
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Reasoning under Uncertainty: Graphical Models

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Conformal prediction is a framework for uncertainty quantification that constructs prediction sets for previously unseen data, guaranteeing coverage of the true label with a specified probability. However, the efficiency of these prediction sets, measured by their size, depends on the choice of the underlying learning model. Relying on a single fixed model may lead to suboptimal performance in online environments, as a single model may not consistently perform well across all time steps. To mitigate this, prior work has explored selecting a model from a set of candidates. However, this approach becomes computationally expensive as the number of candidate models increases. Moreover, poorly performing models in the set may also hinder the effectiveness. To tackle this challenge, this work develops a novel multi-model online conformal prediction algorithm that reduces computational complexity and improves prediction efficiency. At each time step, a bipartite graph is generated to identify a subset of effective models, from which a model is selected to construct the prediction set. Experiments demonstrate that our method outperforms existing multi-model conformal prediction techniques in terms of both prediction set size and computational efficiency.

Problem

• conformal prediction
• online learning
• multi-model selection
• computational efficiency
• prediction set

Innovation

• multi-model conformal prediction
• online learning
• bipartite graph selection
• computational efficiency
• prediction set size