Skip to content

The MATLAB code below implements the second-order SPSA (simultaneous perturbation stochastic approximation) and second-order SG (stochastic gradient) in an efficient manner, such that the FLOPS-per-iteration is reduced from $O(p^3)$ to $O(p^2)$ where $p$ is the dimension of the underlying parameter. See more details in "Efficient Implementation …

License

jingyi-zhu/Fast2SPSA

Repository files navigation

Fast2SPSA

The MATLAB code below implements the second-order SPSA (simultaneous perturbation stochastic approximation) and second-order SG (stochastic gradient) in an efficient manner, such that the FLOPS-per-iteration is reduced from $O(p^3)$ to $O(p^2)$ where $p$ is the dimension of the underlying parameter.

Remark

If you find it useful, please consider citing our IEEE TNNLS article link. Thanks!

About

The MATLAB code below implements the second-order SPSA (simultaneous perturbation stochastic approximation) and second-order SG (stochastic gradient) in an efficient manner, such that the FLOPS-per-iteration is reduced from $O(p^3)$ to $O(p^2)$ where $p$ is the dimension of the underlying parameter. See more details in "Efficient Implementation …

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages