US 11,836,665 B2
Explainable process prediction
Roeland Johannus Scheepens, Eindhoven (NL); and Celine Verhoef, Eindhoven (NL)
Assigned to UiPath, Inc., New York, NY (US)
Filed by UiPath, Inc., New York, NY (US)
Filed on Dec. 30, 2019, as Appl. No. 16/729,971.
Prior Publication US 2021/0201184 A1, Jul. 1, 2021
Int. Cl. G06Q 10/00 (2023.01); G06Q 10/0637 (2023.01); G06N 20/00 (2019.01); G06N 20/20 (2019.01)
CPC G06Q 10/06375 (2013.01) [G06N 20/00 (2019.01); G06N 20/20 (2019.01)] 22 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
generating predictions, using a particular machine learning-based prediction model, for one or more process parameters associated with a running process by selecting the particular machine learning-based prediction model from among one or more trained machine learning-based prediction models for generating the predictions in an on-line mode for the running process, the particular machine learning-based prediction model receiving as input features extracted from prefix traces of an event log associated with execution of the running process and generating the predictions, wherein the one or more trained machine learning-based prediction models are trained in an off-line mode using a respective training set including closed cases;
generating explanation-oriented data elements corresponding to the generated predictions, the explanation-oriented data elements including confidence indicators associated with the generated predictions; and
presenting the explanation-oriented data elements in one or more dashboards of a visualization platform.