Trading off Complexity With Communication Costs in Distributed Adaptive Learning via Krylov Subspaces for Dimensionality Reduction.
Symeon ChouvardasKonstantinos SlavakisSergios TheodoridisPublished in: IEEE J. Sel. Top. Signal Process. (2013)
Keyphrases
- communication cost
- adaptive learning
- dimensionality reduction
- network size
- high dimensional data
- communication overhead
- data distribution
- sensor networks
- principal component analysis
- distributed data
- low dimensional
- lower dimensional
- data points
- high dimensional
- reduce communication cost
- learning environment
- feature space
- linear projection
- learning objects
- concept maps
- reinforcement learning
- learning resources
- databases
- processing cost
- data sets
- data streams
- e learning
- distributed systems
- management system
- knn
- feature selection