Linear Convergence of Gradient and Proximal-Gradient Methods Under the Polyak-Łojasiewicz Condition.
Hamed KarimiJulie NutiniMark SchmidtPublished in: ECML/PKDD (1) (2016)
Keyphrases
- machine learning
- gradient method
- significant improvement
- computational cost
- machine learning methods
- neural network
- image segmentation
- quasi newton
- optimization methods
- machine learning algorithms
- benchmark datasets
- statistical methods
- steepest ascent
- methods require
- empirical studies
- edge detection
- evolutionary algorithm
- feature vectors
- learning algorithm