New ApolloPFN model improves time series forecasting with exogenous variables

Researchers have developed ApolloPFN, a novel time-aware Prior Fitted Network designed to improve zero-shot forecasting by incorporating exogenous variables. Unlike existing foundation models that rely solely on historical time series data, ApolloPFN integrates external factors like promotions, prices, or temperature to enhance accuracy. The system features a synthetic data generation framework for realistic temporal patterns and architectural modifications to leverage temporal context, outperforming current baselines on benchmarks including M5 and solar energy datasets. AI

IMPACT Enhances forecasting accuracy by incorporating external factors, potentially improving applications in retail, energy, and traffic prediction.

RANK_REASON This is a research paper detailing a new model for time series forecasting. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• ApolloPFN
• Chronos
• LagLlama
• Ravi Kiran Selvam
• solar energy
• sundial
• TimeLLM
• TimeMoE
• TimesFM
• UCI Air Quality

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →