Towards the full information chain theory: solution methods for optimal information acquisition problem

TL;DR

This paper develops solution methods for the optimal information acquisition problem in stochastic decision making, linking information source accuracy and relevance, and proposing approximation techniques using probability metrics.

Contribution

It introduces a framework connecting information accuracy and relevance, and develops approximate solution methods for optimal information acquisition in stochastic optimization.

Findings

Proposed a framework relating information accuracy to relevance.

Developed approximate solution methods using probability metrics.

Applied scenario reduction techniques in stochastic optimization.

Abstract

When additional information sources are available in decision making problems that allow stochastic optimization formulations, an important question is how to optimally use the information the sources are capable of providing. A framework that relates information accuracy determined by the source's knowledge structure to its relevance determined by the problem being solved was proposed in a companion paper. There, the problem of optimal information acquisition was formulated as that of minimization of the expected loss of the solution subject to constraints dictated by the information source knowledge structure and depth. Approximate solution methods for this problem are developed making use of probability metrics method and its application for scenario reduction in stochastic optimization.