Skip to content

Ak2714/WorkloadDiff

Repository files navigation

WorkloadDiff (TCC 2024)

The repository is the official implementation for the paper: Weiping Zheng, Zongxiao Chen, Kaiyuan Zheng, Weijian Zheng, Yiqi Chen, Xiaomao Fan, "WorkloadDiff: Conditional Denoising Diffusion Probabilistic Models for Cloud Workload Prediction", IEEE Transactions on Cloud Computing.

We have devised a two-path neural network tailored for cloud workload prediction. This network consists of four residual layers, with bidirectional dilated convolution (Bi-DilConv) serving as the fundamental building block.

Prerequisites

Dependencies can be installed using the following command:

pip install -r requirements

Usage

Training and forecasting for the workload dataset.

python exe_main.py --nsample [number of samples]

Execute multiple times and calculate the average.

sh exe_WorkloadDiff.sh

Make predictions using a pre-trained model.

python exe_main.py --modelfolder [model file path]

Acknowledgement

We are grateful for the valuable codes in the following GitHub repositories, which have been instrumental in achieving our baseline results.

Citation

@article{zheng2024workloaddiff,
  title={WorkloadDiff: Conditional Denoising Diffusion Probabilistic Models for Cloud Workload Prediction},
  author={Zheng, Weiping and Chen, Zongxiao and Zheng, Kaiyuan and Zheng, Weijian and Chen, Yiqi and Fan, Xiaomao},
  journal={IEEE Transactions on Cloud Computing},
  year={2024},
  publisher={IEEE}
}

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published