Federated Learning Model Training Method Based on Data Features Perception Aggregation.
Zeng YanYan Zhong YiJilin ZhangNailiang ZhaoYongjian RenJian WanJun YuPublished in: VTC Fall (2021)
Keyphrases
- prior knowledge
- learned models
- input data
- test data
- synthetic data
- learning models
- prior information
- classification method
- statistical methods
- data sets
- experimental data
- learning scheme
- statistical information
- training process
- bayesian methods
- noisy data
- training data
- learning algorithm
- em algorithm
- predictive model
- mathematical model
- automatically learned
- background knowledge
- dimensional vector
- generative model
- data sources
- objective function
- probabilistic model
- training samples
- statistical model
- classification accuracy
- supervised learning
- sensory data
- confidence scores
- data points
- supervised training
- extracted features
- feature set
- missing data
- training examples
- parameter estimation
- latent factors
- structured prediction
- bayesian networks
- probability distribution
- missing values
- multilayer neural network
- unsupervised learning
- log linear models
- reinforcement learning
- neural network
- weight vector
- object detectors
- training set
- learning process
- active learning
- hidden variables
- training algorithm
- support vector machine
- model selection