Revisiting Gradient Clipping: Stochastic bias and tight convergence guarantees.
Anastasia KoloskovaHadrien HendrikxSebastian U. StichPublished in: CoRR (2023)
Keyphrases
- stochastic approximation
- convergence proof
- convergence rate
- edge detection
- lower bound
- upper bound
- gradient method
- convergence speed
- monte carlo
- gradient information
- stochastic optimization
- worst case
- database
- step size
- image processing
- machine learning
- denoising
- optical flow
- evolutionary algorithm
- faster convergence
- theoretical guarantees
- stochastic programming
- information systems
- stochastic models
- neural network