Skip to content
#

edge-caching

Here are 14 public repositories matching this topic...

A single deep RL agent (Double DQN, PyTorch) that jointly recommends content and manages edge caches. It beats an RL recommender with a fixed caching rule on every seed: +31.1 pt cache-hit rate on one edge (MovieLens 25M, 10/10 seeds) and +35.8 pt across 8 cooperating caches (5/5 seeds). Follow-on to my diploma thesis (TU Crete).

  • Updated Oct 3, 2026
  • Jupyter Notebook

Add this topic to your repo

To associate your repository with the edge-caching topic, visit your repo's landing page and select "manage topics."

Learn more