An adaptive robust defending algorithm against backdoor attacks in federated learning.
Yongkang WangDi-Hua ZhaiYongping HeYuanqing XiaPublished in: Future Gener. Comput. Syst. (2023)
Keyphrases
- learning algorithm
- randomly generated
- cost function
- expectation maximization
- computationally efficient
- detection algorithm
- parameter tuning
- optimization algorithm
- learning speed
- neural network
- convergence rate
- segmentation algorithm
- particle swarm optimization
- probabilistic model
- computational cost
- experimental evaluation
- learning process
- preprocessing
- objective function
- reinforcement learning
- digital images
- high accuracy
- dynamic programming
- search space
- training set
- matching algorithm
- support vector
- learning environment
- learning rules
- learned models