Algorithms with Predictions
Data Structures and Algorithms
2020-06-17 v1
Abstract
We introduce algorithms that use predictions from machine learning applied to the input to circumvent worst-case analysis. We aim for algorithms that have near optimal performance when these predictions are good, but recover the prediction-less worst case behavior when the predictions have large errors.
Cite
@article{arxiv.2006.09123,
title = {Algorithms with Predictions},
author = {Michael Mitzenmacher and Sergei Vassilvitskii},
journal= {arXiv preprint arXiv:2006.09123},
year = {2020}
}
Comments
survey is to appear as a chapter in Beyond the Worst-Case Analysis of Algorithms, a collection edited by Tim Roughgarden. We hope to occasionally update the survey here, with new versions that include discussions of new results and advances in the area of Algorithms with Predictions