QUOTIENT: Two-Party Secure Neural Network Training and Prediction
arXiv:1907.03372
Abstract
Recently, there has been a wealth of effort devoted to the design of secure protocols for machine learning tasks. Much of this is aimed at enabling secure prediction from highly-accurate Deep Neural Networks (DNNs). However, as DNNs are trained on data, a key question is how such models can be also trained securely. The few prior works on secure DNN training have focused either on designing custom protocols for existing training algorithms, or on developing tailored training algorithms and then applying generic secure protocols. In this work, we investigate the advantages of designing training algorithms alongside a novel secure protocol, incorporating optimizations on both fronts. We present QUOTIENT, a new method for discretized training of DNNs, along with a customized secure two-party protocol for it. QUOTIENT incorporates key components of state-of-the-art DNN training such as layer normalization and adaptive gradient methods, and improves upon the state-of-the-art in DNN training in two-party computation. Compared to prior work, we obtain an improvement of 50X in WAN time and 6% in absolute accuracy.
References in corpus (10)
- Adam: A Method for Stochastic Optimization
- On the Convergence of Adam and Beyond
- QSGD: Communication-Efficient SGD via Gradient Quantization and Encoding
- Binarized Neural Networks
- Convolutional Neural Networks using Logarithmic Data Representation
- Mobile Sensor Data Anonymization
- Bitwise Neural Networks
- XONN: XNOR-based Oblivious Deep Neural Network Inference
- High-Accuracy Low-Precision Training
- Understanding Batch Normalization
Cited by in corpus (7)
- Privacy in Deep Learning: A Survey
- Label Leakage and Protection in Two-party Split Learning
- Fusion: Efficient and Secure Inference Resilient to Malicious Servers
- MPC-enabled Privacy-Preserving Neural Network Training against Malicious Attack
- Privacy-preserving Decentralized Aggregation for Federated Learning
- Practical Two-party Privacy-preserving Neural Network Based on Secret Sharing
- A Novel Privacy-Preserving Deep Learning Scheme without Using Cryptography Component