Abstract
Lindell and Pinkas (2002) have proposed the idea of using the techniques of secure multi-party computations to generate efficient algorithms for privacy preserving data-mining. In this context Kiltz, Leander and Malone-Lee (2005) have presented a protocol for the secure distributed computation of the mean and related statistics in a twoparty setting. Their protocol achieves constant round ...
Abstract
Lindell and Pinkas (2002) have proposed the idea of using the techniques of secure multi-party computations to generate efficient algorithms for privacy preserving data-mining. In this context Kiltz, Leander and Malone-Lee (2005) have presented a protocol for the secure distributed computation of the mean and related statistics in a twoparty setting. Their protocol achieves constant round complexity. As a novel suggestion we use a Chebyshev expansion to accelerate this protocol. This approach considerably reduces the overhead of the protocol in terms of both computation and communication. The proposed technique can be applied to other protocols in the field of privacy preserving data-mining as well.