Skip to content

Workshop

Forecasting Workshop with sktime

with Franz Kiraly & Tim Fetzner & Aryan Saini & Muhammad Armaghan Shakir

Wednesday 8 July 11:00 – 13:00 Room M5 (18 Seats)

About This Session

sktime is the community governed open source package for AI with time series, covering learning tasks including forecasting, time series classification, anomaly & changepoint detection. It is available under a permissive license and non-profit owned. This workshop will cover a selection of popular topics in AI based forecasting, with a view towards applications in retail, energy, and manufacturing: - introduction to sktime: uni- and multivariate forecasting, exogenous data, overview of forecasting algorithms - probabilistic forecasting: range and interval forecasting, distribution forecasting, intermittency, high/low scenarios - hierarchical and global forecasting: reconciliation, composite models, cold start problem - deep learning and forecasting foundation models

Topics

  • Open Source
  • Python