Privacy-Preserving cloud-Aided broad learning system

Broad Learning System (BLS) is a new deep learning model proposed recently, which shows its effectiveness in many fields, such as image recognition and fault detection. In this paper, we propose secure, efficient, verifiable outsourcing algorithm for BLS. This enables resource constrained devices to outsource BLS untrusted cloud server complete training, of great significance the promotion application algorithm. Compared with original algorithm, not only improves efficiency on client, but also ensures that sensitive information client will be leaked server. addition, our can verify correctness returned results probability almost 1. Finally, analyze security theory prove algorithms feasibility through experiments.