Awesome

PIDOptimizer (Proportional–Integral–Derivative Optimizer)

This repository contains source code of the CVPR 2018 paper:

A PID Controller Approach for Stochastic Optimization of Deep Networks, Wangpeng An, Haoqian Wang, Qingyun Sun, Jun Xu, Qionghai Dai, Lei Zhang.

Prerequisite:

matplotlib==2.0.2

Train MLP on MNIST DATAST

python mnist_pid.py python mnist_momentum.py python compare.py

<div align="center"> <img src="moment_vs_pid.jpg" width="700px" /> <p>PID Vs. SGD-Momentum</p> </div>

Citation:

If PIDOptimizer is used in your paper/experiments, please cite the following paper.

@InProceedings{An_2018_CVPR,
author = {An, Wangpeng and Wang, Haoqian and Sun, Qingyun and Xu, Jun and Dai, Qionghai and Zhang, Lei},
title = {A PID Controller Approach for Stochastic Optimization of Deep Networks},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2018}
}