SADAM: Stochastic Adam, A Stochastic Operator for First-Order Gradient-based Optimizer
This work addresses optimization efficiency for machine learning practitioners, but it appears incremental as it builds on existing first-order methods like Adam.
The authors tackled the problem of escaping stationary and saddle points in gradient-based optimization by proposing SADAM, a stochastic operator that improves target accuracy without requiring batches or sampling. Validation on biomedical signal decomposition showed that the strategy can be generalized to first-order optimizers and efficiently enhances accuracy.
In this work, to efficiently help escape the stationary and saddle points, we propose, analyze, and generalize a stochastic strategy performed as an operator for a first-order gradient descent algorithm in order to increase the target accuracy and reduce time consumption. Unlike existing algorithms, the proposed stochastic the strategy does not require any batches and sampling techniques, enabling efficient implementation and maintaining the initial first-order optimizer's convergence rate, but provides an incomparable improvement of target accuracy when optimizing the target functions. In short, the proposed strategy is generalized, applied to Adam, and validated via the decomposition of biomedical signals using Deep Matrix Fitting and another four peer optimizers. The validation results show that the proposed random strategy can be easily generalized for first-order optimizers and efficiently improve the target accuracy.