https://github.com/adamdad/filter-gradient-decent

In this paper, we propose Filter Gradient Decent (FGD), an efficient stochastic optimization algorithm that makes a consistent estimation of the local gradient by solving an adaptive filtering problem with different designs of filters.
https://github.com/adamdad/filter-gradient-decent

filters stochastic-gradient-descent variance-reduction

Last synced: about 1 year ago
JSON representation

In this paper, we propose Filter Gradient Decent (FGD), an efficient stochastic optimization algorithm that makes a consistent estimation of the local gradient by solving an adaptive filtering problem with different designs of filters.

• Host: GitHub
• URL: https://github.com/adamdad/filter-gradient-decent
• Owner: Adamdad
• Created: 2020-12-21T01:42:38.000Z (over 5 years ago)
• Default Branch: main
• Last Pushed: 2021-05-18T08:01:43.000Z (about 5 years ago)
• Last Synced: 2025-03-28T12:38:56.761Z (over 1 year ago)
• Topics: filters, stochastic-gradient-descent, variance-reduction
• Language: Python
• Homepage:
• Size: 1.22 MB
• Stars: 11
• Watchers: 2
• Forks: 2
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Filter-Gradient-Decent

Update: This project also include the code for paper

**Kalman Optimizer for Consistent Gradient Descent**

*Xingyi Yang, (ICASSP2021)* [paper](https://ieeexplore.ieee.org/document/9414588)

Course project for ECE 251C UCSD. Code for paper,

**Stochastic Gradient Variance Reduction by Solving a Filtering Problem**

In this paper, we propose Filter Gradient Decent (FGD), an

efficient stochastic optimization algorithm that make consistent

estimation of the local gradient by solving an adaptive filtering

problem with different design of filters.

![](paper/illustration.jpg)

## Usage

- To do later