Park: An Open Platform for Learning-Augmented Computer Systems
Abstract
We present Park, a platform for researchers to experiment with Reinforcement Learning (RL) for computer systems. Using RL for improving the performance of systems has a lot of potential, but is also in many ways very different from, for example, using RL for games. Thus, in this work we first discuss the unique challenges RL for systems has, and then propose Park an open extensible platform, which makes it easier for ML researchers to work on systems problems. Currently, Park consists of 12 real world system-centric optimization problems with one common easy to use interface. Finally, we present the performance of existing RL approaches over those 12 problems and outline potential areas of future work.
Cite
Text
Mao et al. "Park: An Open Platform for Learning-Augmented Computer Systems." Neural Information Processing Systems, 2019.Markdown
[Mao et al. "Park: An Open Platform for Learning-Augmented Computer Systems." Neural Information Processing Systems, 2019.](https://mlanthology.org/neurips/2019/mao2019neurips-park/)BibTeX
@inproceedings{mao2019neurips-park,
title = {{Park: An Open Platform for Learning-Augmented Computer Systems}},
author = {Mao, Hongzi and Negi, Parimarjan and Narayan, Akshay and Wang, Hanrui and Yang, Jiacheng and Wang, Haonan and Marcus, Ryan and Addanki, Ravichandra and Shirkoohi, Mehrdad Khani and He, Songtao and Nathan, Vikram and Cangialosi, Frank and Venkatakrishnan, Shaileshh and Weng, Wei-Hung and Han, Song and Kraska, Tim and Alizadeh, Dr.Mohammad},
booktitle = {Neural Information Processing Systems},
year = {2019},
pages = {2494-2506},
url = {https://mlanthology.org/neurips/2019/mao2019neurips-park/}
}