Quantum federated learning through blind quantum computing

Private distributed learning studies the problem of how multiple entities collaboratively train a shared deep network with their private data unrevealed. With security provided by protocols blind quantum computation, cooperation between physics and machine may lead to unparalleled prospect for solving tasks. In this paper, we introduce protocol that is able utilize computational power remote servers while keeping safe. For concreteness, first single-party delegated training variational classifiers based on computing then extend multiparty incorporated differential privacy. We carry out extensive numerical simulations different real-life datasets encoding strategies benchmark effectiveness our protocol. find robust experimental imperfections secure under gradient attack after incorporation Our results show potential handling computationally expensive tasks privacy guarantees, thus providing valuable guide exploring advantages from perspective in field applications.