Editing Cache replacement policies (section)

=== {{anchor|Cache replacement policies using machine learning}}Machine-learning policies ===
A number of policies have attempted to use [[perceptron]]s, [[markov chain]]s or other types of [[machine learning]] to predict which line to evict.<ref>{{Cite journal |last1=Liu |first1=Evan |last2=Hashemi |first2=Milad |last3=Swersky |first3=Kevin |last4=Ranganathan |first4=Parthasarathy |last5=Ahn |first5=Junwhan |date=2020-11-21 |title=An Imitation Learning Approach for Cache Replacement |url=https://proceedings.mlr.press/v119/liu20f.html |journal=International Conference on Machine Learning |language=en |publisher=PMLR |pages=6237–6247|arxiv=2006.16239 }}</ref><ref>{{Cite book |last1=Jiménez |first1=Daniel A. |last2=Teran |first2=Elvira |title=Proceedings of the 50th Annual IEEE/ACM International Symposium on Microarchitecture |chapter=Multiperspective reuse prediction |date=2017-10-14 |chapter-url=http://dx.doi.org/10.1145/3123939.3123942 |pages=436–448 |location=New York, NY, USA |publisher=ACM |doi=10.1145/3123939.3123942|isbn=9781450349529 |s2cid=1811177 }}</ref> [[Learning augmented algorithm]]s also exist for cache replacement.
<ref name="LykourisVassilvitskii2021">{{cite journal | last1 = Lykouris | first1 = Thodoris | last2 = Vassilvitskii | first2 = Sergei | title = Competitive Caching with Machine Learned Advice | journal = Journal of the ACM | date = 7 July 2021 | volume = 68 | issue = 4 | pages = 1–25 | issn = 0004-5411 | eissn = 1557-735X | doi = 10.1145/3447579 | pmid = | s2cid = 3625405 | url = | doi-access = free | arxiv = 1802.05399 }}</ref><ref name="MitzenmacherVassilvitskii2020">{{cite book | title = Beyond the Worst-Case Analysis of Algorithms | last1 = Mitzenmacher | author-link1 = Michael Mitzenmacher | first1 = Michael | last2 = Vassilvitskii | first2 = Sergei | chapter = Algorithms with Predictions | date = 31 December 2020 | pages = 646–662 | publisher = Cambridge University Press | doi = 10.1017/9781108637435.037 | arxiv=2006.09123 | isbn = 9781108637435 }}</ref>