Nonconvex sampling with the Metropolis-adjusted Langevin algorithm.
Oren MangoubiNisheeth K. VishnoiPublished in: CoRR (2019)
Keyphrases
- cost function
- experimental evaluation
- learning algorithm
- objective function
- preprocessing
- markov chain monte carlo
- computational complexity
- high accuracy
- sampling algorithm
- convergence rate
- times faster
- monte carlo
- detection algorithm
- simulated annealing
- dynamic programming
- optimal solution
- np hard
- theoretical analysis
- parameter estimation
- optimization algorithm
- neural network
- data sets
- worst case
- recognition algorithm
- image segmentation
- bayesian framework
- matching algorithm
- data structure
- energy function
- significant improvement
- model selection