VFDV-IM: An Efficient and Securely Vertical Federated Data Valuation

Vertical federated learning enables multiple participants to build a joint machine learning model upon distributed features of overlapping samples. The performance of VFL models heavily depends on the quality of participants’ local data. It’s essential to measure the contributions of the participants for various purposes, e.g., participant selection and reward allocation. The Shapley value is widely adopted by previous works for contribution assessment. However, computing the Shapley value in VFL requires repetitive model training from scratch, incurring expensive computation and communication overheads. Inspired by this challenge, in this paper, we ask: can we efficiently and securely perform data valuation for participants via the Shapley value in VFL? We call this problem Vertical Federated Data Valuation, and introduce VFDV-IM, a method utilizing an Inheritance Mechanism to expedite Shapley value calculations by leveraging historical training records. We first propose a simple, yet effective, strategy that directly inherits the model trained over the entire consortium. To further optimize VFDV-IM, we propose a model ensemble approach that measures the similarity of evaluated consortiums, based on which we reweight the historical models. We conduct extensive experiments on various datasets and show that our VFDV-IM can efficiently calculate the Shapley value while maintaining accuracy.