Generative Forecasting: The Power of Graph Transformers in Time Series Prediction скачать в хорошем качестве

Generative Forecasting: The Power of Graph Transformers in Time Series Prediction

Time series forecasting is essential for anticipating trends, behaviors, and business outcomes. Yet, traditional methods often miss the rich context embedded in interconnected data. In this online workshop, Jan Eric Lenssen and Sally Liu from Kumo will explore how Graph Transformers, powered by relational data, are reshaping the way enterprises forecast. We’ll walk through an end-to-end forecasting pipeline built with Graph Transformers, showing how to generate more accurate predictions by combining temporal, relational, and contextual signals. You’ll also learn the difference between predictive and generative approaches, see real-world use cases, and understand how this method improves performance across forecasting tasks.