Perturbed Projected Gradient Descent Converges to Approximate Second-order Points for Bound Constrained Nonconvex Problems.
Songtao LuZiping ZhaoKejun HuangMingyi HongPublished in: ICASSP (2019)
Keyphrases
- objective function
- optimization problems
- lower bound
- np complete
- convex functions
- solving problems
- exact solution
- global optimization
- point sets
- higher order
- simulated annealing
- feature points
- evolutionary algorithm
- search algorithm
- convex optimization
- computer vision
- neural network
- variational inequalities
- data sets