Feedforward Neural Networks for Caching: Enough or Too Much?

Abstract : We propose a caching policy that uses a feedforward neural network (FNN) to predict content popularity. Our scheme outperforms popular eviction policies like LRU or ARC, but also a new policy relying on the more complex recurrent neural networks. At the same time, replacing the FNN predictor with a naive linear estimator does not degrade caching performance significantly, questioning then the role of neural networks for these applications.
Complete list of metadatas

Cited literature [18 references]  Display  Hide  Download

https://hal.inria.fr/hal-02411461
Contributor : Giovanni Neglia <>
Submitted on : Saturday, December 14, 2019 - 8:39:45 PM
Last modification on : Tuesday, December 17, 2019 - 2:06:23 AM

File

1810.06930(1).pdf
Files produced by the author(s)

Identifiers

Collections

Citation

Vladyslav Fedchenko, Giovanni Neglia, Bruno Ribeiro. Feedforward Neural Networks for Caching: Enough or Too Much?. ACM SIGMETRICS Performance Evaluation Review, Association for Computing Machinery, 2019, 46 (3), pp.139-142. ⟨10.1145/3308897.3308958⟩. ⟨hal-02411461⟩

Share

Metrics

Record views

10

Files downloads

47