[en] In-line scheduling schemes based on contextual bandits. The performance of each flavor of bandits (Upper confidence band (UCB) and KL-UCB as well as contextual variants such as time-limited k-nearest neighbors contextual UCB which displayed the best results given some regularity in the reward function (possibly less than Lipfschitz).
Research Center/Unit :
MiMeR
Disciplines :
Geological, petroleum & mining engineering
Author, co-author :
Koch, Pierre-Henri ; Université de Liège - ULiège > Département ArGEnCo > Géoressources minérales & Imagerie géologique
Language :
English
Title :
Inline block scheduling optimization for mining operations
Publication date :
June 2019
Creation date :
2019-01
Name of the research project :
PREP
Funders :
VINNOVA [SE]
Commentary :
The code was made available to MiMeR research unit at Lulea Tekniska Universitet in June 2019. Most of the code is in Matlab with some Python.
