OPAF: Optimized Secure Two-Party Computation Protocols for Nonlinear Activation Functions in Recurrent Neural Network.
Qian FengZhihua XiaZhifeng XuJiasi WengJian WengPublished in: CoRR (2024)
Keyphrases
- recurrent neural networks
- activation function
- hidden layer
- complex valued
- feed forward
- nonlinear dynamic systems
- key agreement protocol
- neural network
- artificial neural networks
- feedforward neural networks
- echo state networks
- feed forward neural networks
- neural model
- back propagation
- real valued
- nonlinear functions
- cryptographic protocols
- key distribution
- connection weights
- neuro fuzzy
- recurrent networks
- neural nets
- network architecture
- learning rate
- semi honest
- training data
- diffie hellman
- key management
- multi layer perceptron