The Lazy Online Subgradient Algorithm is Universal on Strongly Convex Domains.
Daron AndersonDouglas J. LeithPublished in: NeurIPS (2021)
Keyphrases
- experimental evaluation
- learning algorithm
- detection algorithm
- significant improvement
- matching algorithm
- optimization algorithm
- expectation maximization
- simulated annealing
- computational cost
- dynamic programming
- np hard
- piecewise linear
- k means
- search space
- convergence rate
- times faster
- globally optimal
- optimization procedure
- passive aggressive
- particle swarm optimization
- high accuracy
- cost function
- preprocessing
- optimal solution
- kolmogorov complexity
- similarity measure