On orthogonal projections for dimension reduction and applications in variational loss functions for learning problems.
Anna BregerJosé Ignacio OrlandoPavol HarárMonika DörflerSophie KlimschaChristoph GrechenigBianca S. GerendasUrsula Schmidt-ErfurthMartin EhlerPublished in: CoRR (2019)
Keyphrases
- dimension reduction
- learning problems
- loss function
- learning tasks
- principal component analysis
- supervised learning
- machine learning algorithms
- kernel methods
- feature extraction
- support vector
- high dimensional
- pairwise
- learning algorithm
- binary classification
- random projections
- reproducing kernel hilbert space
- linear discriminant analysis
- machine learning
- learning models
- singular value decomposition
- feature space
- low dimensional
- dimensionality reduction
- image segmentation
- semi supervised learning
- unsupervised learning
- cluster analysis
- reinforcement learning
- high dimensional data
- multi task
- feature selection
- data sets
- image classification
- semi supervised
- high dimensionality
- multiple kernel learning