Intrinsic persistent homology via density-based metric learning.
Eugenio BorghiniXimena FernándezPablo GroismanGabriel B. MindlinPublished in: CoRR (2020)
Keyphrases
- metric learning
- persistent homology
- topological features
- distance metric learning
- distance metric
- morse theory
- pairwise
- dimensionality reduction
- distance function
- learning tasks
- semi supervised
- computational geometry
- clustering algorithm
- multi task
- semi supervised learning
- mahalanobis metric
- feature space
- labeled data
- unsupervised learning
- text classification
- background knowledge
- geometric structure
- transfer learning
- euclidean distance
- data mining
- learning experience
- computer vision