Get Started
Home
Topics
Search
Library
Research questionHow should temporal cascade prediction benchmarks avoid leakage while reflecting downstream conversion outcomes?Randomly splitting cascades can let future signals enter evaluation, inflating apparent predictive performance. Benchmarks that measure popularity without observed conversions may also fail to reflect whether predicted cascades lead to downstream purchases.
AI
Evaluation & Benchmarks
Machine Learning
Statistical Machine Learning
Latest papersRecent research connected to this question, newest first.From Leakage to Fidelity: Reliable Benchmarking for Temporal Cascade PredictionThe source presents a Full Temporal protocol, overlap-based leakage diagnostics, and analyses of performance inflation and temporal drift. Its Taoke e-commerce dataset supplies promoter and product features plus observed purchase conversions for popularity and conversion forecasting; the evidence also includes a lightweight reference pipeline and a larger same-task internal extension. Findings are grounded in temporal cascade benchmarking and this e-commerce setting.research paper · Sep 3, 2026
Related questions
How can PII detectors avoid missed entities when deployment data shifts from benchmark conditions?Can phase-transition counts during fine-tuning predict final test accuracy across architectures and distribution shifts?How can synthetic medical time-series generation preserve rare-class patterns across temporal scales for downstream prediction?How can we reduce per-task LLM-agent evaluation cost without distorting benchmark outcomes?