Adaptive Elicitation of Preferences under Uncertainty in Sequential Decision Making Problems

Authors: Nawal Benabbou, Patrice Perny

IJCAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We propose an interactive algorithm with performance guarantees and describe numerical tests demonstrating the practical efficiency of our approach. and Finally, we introduce a representation of imprecise utilities using spline functions and we provide numerical tests showing the practical efficiency of the proposed approach (Section 4). and Moreover, we report the results of numerical tests aiming to evaluate the performance of our interactive search procedure (Algorithm 3) on randomly generated instances of perfect binary trees (alternating decision nodes and chance nodes along each path).
Researcher Affiliation Academia Nawal Benabbou and Patrice Perny Sorbonne Universit es, UPMC Univ Paris 06, UMR 7606, LIP6 CNRS, UMR 7606, LIP6, F-75005, Paris, France
Pseudocode Yes Algorithm 1: Backward Induction., Algorithm 2: Combination., Algorithm 3: Interactive Backward Induction.
Open Source Code No The paper does not provide any concrete access information (e.g., specific repository link, explicit code release statement) for source code.
Open Datasets No Moreover, we report the results of numerical tests aiming to evaluate the performance of our interactive search procedure (Algorithm 3) on randomly generated instances of perfect binary trees (alternating decision nodes and chance nodes along each path).
Dataset Splits No The paper does not specify exact split percentages, sample counts, citations to predefined splits, or detailed splitting methodology for dataset partitioning.
Hardware Specification Yes The tests are performed on a Intel Core i7-4770 CPU with 15GB of RAM.
Software Dependencies No LPs are optimized using the Gurobi solver.
Experiment Setup Yes The depth d of the tree varies from 8 to 18 and the tolerance threshold from δ = 0.05 to 0.1. and Moreover, we have f(ℓλ, u) = λu(x ) + (1 λ)u(x ) = λ for all λ [0, 1]. and We can impose u(x ) = 0 and u(x ) = 1 since v NM utilities are unique up to positive affine transformations. and In this paper, we use I-spline functions of order 3 with a uniform subdivision of the unit interval (assuming that outcomes have been normalized to belong to the unit interval).