Announcing MCP Toolbox Java SDK v1.0: Agentic data access for the enterprise
cloud.google.com·6d ago
Cloud & Infrastructure
The New Stack·Aug 31·5 min read
Google’s new forecasting model beats everyone. You can’t use it at work (yet).
TL;DR
Google has introduced TimesFM-3, a groundbreaking time-series forecasting model with 330 million parameters, outperforming existing models. However, it is not yet available for workplace use.
✦ Why It Matters
Stay informed about emerging forecasting technologies to prepare for future integration into your projects.
Key Takeaways
- Experiment with TimesFM-3 for advanced forecasting in research projects.
- Incorporate multiple time series and external factors for better predictions.
- Stay updated on licensing changes for potential commercial use.
- Benchmark your existing models against TimesFM-3 for performance insights.
How It Works
TimesFM-3 employs a decoder-only transformer architecture that processes time series data in patches. It uses two types of attention layers: one that looks backward in time within a single series and another that examines multiple series simultaneously.
This dual attention mechanism allows the model to make more informed forecasts by considering related data points and external influences.