A HistGradientBoosting model predicts pre-execution BigQuery slot-time from query complexity, volume, and text features, cutting error 30-37% versus baselines on significant queries while failing on long-tail outliers.
Towards multi-tenant performance SLOs
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.DB 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
HSSPS uses heuristics and client-side tokens to partition queries, delivering 50-97% P95 latency reductions and 8-10x throughput gains in large cloud databases.
citing papers explorer
-
Pre-Execution Query Slot-Time Prediction in Cloud Data Warehouses: A Feature-Scoped Machine Learning Approach
A HistGradientBoosting model predicts pre-execution BigQuery slot-time from query complexity, volume, and text features, cutting error 30-37% versus baselines on significant queries while failing on long-tail outliers.
-
Heuristic Search Space Partitioning for Low-Latency Multi-Tenant Cloud Queries
HSSPS uses heuristics and client-side tokens to partition queries, delivering 50-97% P95 latency reductions and 8-10x throughput gains in large cloud databases.