Collaborative Prediction: Tractable Information Aggregation via Agreement

2025-04-08

AI Summary

This paper addresses two-party collaborative prediction under feature partitioning: parties exchange only label predictions—neither raw features nor shared priors are permitted. We propose a communication-efficient, computationally feasible non-Bayesian collaborative protocol that generalizes Aumann’s agreement theorem to a statistically robust framework. Our method integrates online adversarial learning with batch statistical learning, enabling prediction reduction via coordinated communication between local learners and best-response actions. Crucially, communication complexity is independent of data dimensionality, and the protocol achieves provably optimal collaborative performance. Theoretical analysis establishes tight regret bounds for the online setting and yields practical algorithms for the batch setting, both guaranteeing strong generalization and computational tractability. The approach is particularly suited for privacy-sensitive applications such as human-AI interaction and multimodal learning.

Technology Category

• Game Theory and Economic Paradigms: Adversarial Learning
• Machine Learning: Online Learning & Bandits
• Multiagent Systems: Adversarial Agents

Application Category

• Economics, Online Markets and Human Computation: Social networks and social learning
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems

Abstract

We give efficient"collaboration protocols"through which two parties, who observe different features about the same instances, can interact to arrive at predictions that are more accurate than either could have obtained on their own. The parties only need to iteratively share and update their own label predictions-without either party ever having to share the actual features that they observe. Our protocols are efficient reductions to the problem of learning on each party's feature space alone, and so can be used even in settings in which each party's feature space is illegible to the other-which arises in models of human/AI interaction and in multi-modal learning. The communication requirements of our protocols are independent of the dimensionality of the data. In an online adversarial setting we show how to give regret bounds on the predictions that the parties arrive at with respect to a class of benchmark policies defined on the joint feature space of the two parties, despite the fact that neither party has access to this joint feature space. We also give simpler algorithms for the same task in the batch setting in which we assume that there is a fixed but unknown data distribution. We generalize our protocols to a decision theoretic setting with high dimensional outcome spaces, where parties communicate only"best response actions."Our theorems give a computationally and statistically tractable generalization of past work on information aggregation amongst Bayesians who share a common and correct prior, as part of a literature studying"agreement"in the style of Aumann's agreement theorem. Our results require no knowledge of (or even the existence of) a prior distribution and are computationally efficient. Nevertheless we show how to lift our theorems back to this classical Bayesian setting, and in doing so, give new information aggregation theorems for Bayesian agreement.

Problem

• Efficient collaboration protocols for accurate joint predictions without sharing features
• Regret bounds for adversarial online learning without joint feature access
• Generalization of Bayesian agreement theorems without prior distribution knowledge

Innovation

• Efficient collaboration protocols without feature sharing
• Communication independent of data dimensionality
• Generalization to high dimensional outcome spaces