FedSA: A staleness-aware asynchronous Federated Learning algorithm with non-IID data.
Ming ChenBingcheng MaoTianyi MaPublished in: Future Gener. Comput. Syst. (2021)
Keyphrases
- data sources
- learning algorithm
- data analysis
- data sets
- original data
- distributed data
- high quality
- knowledge discovery
- small number
- data mining
- noisy data
- raw data
- synthetic data
- training examples
- data mining techniques
- data processing
- data collection
- statistical analysis
- image data
- data points
- active learning
- data quality
- historical data
- training data
- computer systems
- training samples
- high dimensional data
- distributed systems
- sensor data
- data objects