pipette
ENEnglish

AdaptLSTM: Efficient Adaptive Online Learning for Cloud Workload Forecasting under Distribution Drift

Xinhua Miao, Bowei Yang, Zhengong Cai

PreprintUso en el mundo real

En palabras de los autores

Accurate workload forecasting is critical for elastic resource provisioning in web-scale cloud services, where distribution shifts driven by viral content, product launches, and user behavior degrade offline-trained models rapidly. Naive online learning recovers accuracy but incurs prohibitive per-step compute cost. We propose AdaptLSTM, an adaptive online framework that detects drift via validation-calibrated thresholds and applies selective, targeted updates. On the Alibaba Machine Trace, AdaptLSTM recovers 54% of Naive Online's improvement at 20% cost ( efficiency, over 10 seeds). On the more volatile Container Trace, it achieves 96% at 20% cost ( efficiency, MAE reduction over Static). Unlike classical drift detectors (ADWIN, DDM, Page-Hinkley) which fail to trigger on regression-scale error streams, AdaptLSTM fires 42 times over 301 steps and outperforms matched-budget baselines. Wall-clock profiling shows throughput gain and 45% update-time reduction. The framework is model-agnostic: identical Pareto patterns hold for LSTM, GRU, and Transformer backbones.

Resultado principalEl resumen no menciona limitaciones.

Apareció: viernes, 9 de octubre. arXiv. Preprint, todavía sin revisión por pares.