Incentive Allocation in Vertical Federated Learning Based on Bankruptcy Problem

2023-07-07

arXiv.org

AI Summary

In vertical federated learning (VFL), passive parties often lack sufficient incentive to participate due to unclear or unfair reward allocation. Method: This paper proposes a bankruptcy-game-based incentive mechanism wherein the active party allocates rewards to passive parties proportionally to their data contributions. It innovatively models the reward allocation problem as a bankruptcy game and adopts the Talmudic rule, solved via the nucleolus to ensure fairness and stability. Compared to the Shapley value, this approach reduces computational complexity by over an order of magnitude. Results: Extensive experiments on synthetic and real-world datasets demonstrate that the mechanism achieves theoretical rigor, fair contribution-based allocation, strategic stability (i.e., resistance to coalition manipulation), and high scalability. It establishes a novel paradigm for incentive design in VFL, bridging cooperative game theory with practical federated learning systems.

Abstract

Vertical federated learning (VFL) is a promising approach for collaboratively training machine learning models using private data partitioned vertically across different parties. Ideally in a VFL setting, the active party (party possessing features of samples with labels) benefits by improving its machine learning model through collaboration with some passive parties (parties possessing additional features of the same samples without labels) in a privacy preserving manner. However, motivating passive parties to participate in VFL can be challenging. In this paper, we focus on the problem of allocating incentives to the passive parties by the active party based on their contributions to the VFL process. We address this by formulating the incentive allocation problem as a bankruptcy game, a concept from cooperative game theory. Using the Talmudic division rule, which leads to the Nucleolus as its solution, we ensure a fair distribution of incentives. We evaluate our proposed method on synthetic and real-world datasets and show that it ensures fairness and stability in incentive allocation among passive parties who contribute their data to the federated model. Additionally, we compare our method to the existing solution of calculating Shapley values and show that our approach provides a more efficient solution with fewer computations.

Problem

• Allocate incentives in Vertical Federated Learning.
• Formulate incentive allocation as a bankruptcy game.
• Ensure fair and stable incentive distribution.

Innovation

• Formulates incentive allocation as bankruptcy game
• Uses Talmudic division rule for fair distribution
• Ensures fairness, stability in VFL incentive allocation