Learn More
This paper develops a distributed dictionary learning algorithm for sparse representation of the data distributed across nodes of sensor networks, where the sensitive or private data are stored or there is no fusion center or there exists a big data application. The main contributions of this paper are: 1) we decouple the combined dictionary atom update and(More)
This paper develops a novel decentralized dimensionality reduction algorithm for the distributed tensor data across sensor networks. The main contributions of this paper are as follows. First, conventional centralized methods, which utilize entire data to simultaneously determine all the vectors of the projection matrix along each tensor mode, are not(More)
  • 1