XMANAI Hybrid model
EU-funded XMANAI project deals with bringing explainable AI to the Industry, and this asset is an example of the models developed during the project.
This model works on the specific case of predicting work line production given information from several points in the line, aggregated over time. Explanations are then performed on the set of predictions.
The provided toy data example contains the production at 14 points in the line (a.k.a. teams) across a set of 8 hours, i.e., a full work turn. Each row of the data include production information for each one of these teams at each hour, plus a target that is the end (average) production of the line for that turn. Data have been preprocessed so each turn also includes the information from the previous turn (columns appended with "_old").