Flexible data representation with graph convolution for semi-supervised learning.
Fadi DornaikaPublished in: Neural Comput. Appl. (2021)
Keyphrases
- data representation
- semi supervised learning
- graph construction
- label propagation
- graph based semi supervised learning
- semi supervised classification
- semi supervised
- unlabeled data
- graph kernels
- dimensionality reduction
- labeled data
- unsupervised learning
- co training
- subspace learning
- supervised learning
- data representations
- pairwise constraints
- machine learning
- active learning
- training data
- semi supervised learning methods
- image processing
- xml documents
- weighted graph
- feature space
- labeled and unlabeled data
- data sets
- knowledge base
- discriminative information
- reinforcement learning
- xml schema
- nonnegative matrix factorization
- structured data
- metric learning
- learning models