Faster proximal algorithms for matrix optimization using Jacobi-based eigenvalue methods.
Hamza FawziHarry GoulbournePublished in: NeurIPS (2021)
Keyphrases
- optimization methods
- significant improvement
- computational cost
- optimization problems
- benchmark datasets
- computationally expensive
- machine learning algorithms
- search methods
- methods outperform
- methods require
- linear algebra
- computationally intensive
- quasi newton
- machine learning methods
- orders of magnitude faster
- data mining techniques
- monte carlo methods
- evolutionary algorithm
- heuristic methods
- efficient optimization
- synthetic and real datasets
- theoretical guarantees
- discrete optimization
- preprocessing
- data mining
- genetic algorithm
- learning algorithm
- missing values
- computer vision algorithms
- constrained optimization
- statistical methods
- optimization method