EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

2385. Optimal Testing Strategies for Information Acquisition in Decision Programming

Invited abstract in session WA-34: Decision problems represented as influence diagrams, stream Stochastic, Robust and Distributionally Robust Optimization.

Wednesday, 8:30-10:00
Room: 43 (building: 303A)

Authors (first author is the speaker)

1. Ahti Salo
Systems Analysis Laboratory, Aalto University School of Science
2. Topias Terho
Department of mathemathics and systems analysis, Aalto University
3. Fabricio Oliveira
Mathematics and Systems Analysis, Aalto University

Abstract

In the context of influence diagrams, we formalise the notion of information decisions which are associated with decision nodes that determine what information will be acquired to support decisions at a given decision node later; or, more specifically, what is the set of nodes whose states are known when making this later decision. Examples of such information decisions include choices among diagnostic tests because testing decisions determine what test results will be available when making treatment decisions, for instance. For the development of optimal strategies, we build on Decision Programming (Salo et al., EJOR 299/2, 2022) in which the influence diagram is solved by converting it into an equivalent mixed-integer linear programming (MILP) problem; this makes it possible to solve influence diagrams which do not fulfil the usual ‘no-forgetting’ assumption, or which involve logical, resource, or risk constraints that preclude the use of common solution approaches such as arc reversals and node removals. We present optimization formulations that employ binary variables to represent information decisions. We also provide numerical examples to illustrate the development of optimal testing strategies for information acquisition.

Keywords

Status: accepted

Back to the list of papers