Secure Two-Party Feature Selection
This addresses privacy concerns for data owners and acquirers in machine learning by enabling secure feature selection, though it is incremental as it builds on existing secure computation frameworks.
The paper tackles the problem of securely evaluating data value for trading in classification tasks without a trusted third party, proposing a four-round protocol that is provably secure in the honest-but-curious adversary model.
In this work, we study how to securely evaluate the value of trading data without requiring a trusted third party. We focus on the important machine learning task of classification. This leads us to propose a provably secure four-round protocol that computes the value of the data to be traded without revealing the data to the potential acquirer. The theoretical results demonstrate a number of important properties of the proposed protocol. In particular, we prove the security of the proposed protocol in the honest-but-curious adversary model.