Griffin uses an interpretable random forest to attribute a slow job's runtime deviation to individual metrics and rank the likely causes without labeled data.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Griffon: Reasoning about Job Anomalies with Unlabeled Data in Cloud-based Platforms
Griffin uses an interpretable random forest to attribute a slow job's runtime deviation to individual metrics and rank the likely causes without labeled data.