LAFC-Evict: Learning-Augmented Cache Eviction Dataset

LAFC-Evict is a derived, tabular research dataset of counterfactual supervision labels for learning-augmented cache-eviction research. Each row describes one candidate object considered for eviction at one full-cache-miss decision point, together with engineered state/candidate features and labels describing what would happen to the cache miss rate over a short future horizon if that specific candidate were evicted. The current AWS release, v1.0, contains five trace families: cloudphysics, metacdn, metakv, twemcache, and wiki2018. It includes 277,995,072 candidate rows, 2,363,286 decision-view...