Bounded Approximations for Linear Multi-Objective Planning Under Uncertainty

Roijers, Diederik M, Scharpff, Joris, Spaan, Matthijs TJ, Oliehoek, Frans A ORCID: 0000-0003-4372-5055, de Weerdt, Mathijs and Whiteson, Shimon
(2014) Bounded Approximations for Linear Multi-Objective Planning Under Uncertainty. .

[img] Text
Roijers14ICAPS.pdf - Unspecified

Download (394kB)


Planning under uncertainty poses a complex problem in which multiple objectives often need to be balanced. When dealing with multiple objectives, it is often assumed that the relative importance of the objectives is known a priori. How-ever, in practice human decision makers often find it hard to specify such preferences, and would prefer a decision support system that presents a range of possible alternatives. We propose two algorithms for computing these alternatives for the case of linearly weighted objectives. First, we propose an anytime method, approximate optimistic linear support (AOLS), that incrementally builds up a complete set of e-optimal plans, exploiting the piecewise-linear and convex shape of the value function. Second, we propose an approx-imate anytime method, scalarised sample incremental im-provement (SSII), that employs weight sampling to focus on the most interesting regions in weight space, as suggested by a prior over preferences. We show empirically that our methods are able to produce (near-)optimal alternative sets orders of magnitude faster than existing techniques.

Item Type: Conference or Workshop Item (Unspecified)
Additional Information: bib2html_pubtype: Refereed Conference (International) bib2html_rescat: Multiobjective Decision Making
Depositing User: Symplectic Admin
Date Deposited: 12 Apr 2016 11:14
Last Modified: 15 Dec 2022 22:07
Related URLs: