Gaussian Approximations of SDES in Metropolis-Adjusted Langevin Algorithms.
Simo SärkkäChristos MerkatasToni KarvonenPublished in: MLSP (2021)
Keyphrases
- theoretical analysis
- orders of magnitude
- computationally efficient
- computational cost
- significant improvement
- learning algorithm
- computational complexity
- data structure
- worst case
- neural network
- reinforcement learning
- particle filter
- image segmentation
- image processing
- benchmark datasets
- data mining algorithms
- computationally expensive