Nonlinear temporal and spatial forecasting: modelling and uncertainty analysis (NoTeS) – MASIT20
Risto Ritala, Esa Alhoniemi, Tuomo Kauranne, Kimmo Konkarikoski, Amaury Lendasse and Miki Sirola
In: MASI Programme 2005–2009,Yearbook 2008 and workshop, 2008, Finland.
NoTeS project is developing a generic toolset for spatio-temporal forecasting and forecast uncertainty
analysis through analyzing five widely different test cases: forecasting energy consumption, diagnosing
quality variations at a pulp mill, estimating and controlling paper quality on the basis scanning sensors,
supporting operational decisions at nuclear power plant, and segmenting medical images. An abstract set
of functions common to all these applications – with possible exception of image segmentation – has
been identified, and the corresponding structures for data, models and estimates specified. The project is
in stage in which the test cases have demonstrated practical relevance and economic potential and in
which the specification work for generic toolset is stabilizing.