Solving Influence Diagrams using HUGIN, Shafer-Shenoy and Lazy Propagation

TL;DR

This paper compares three architectures for influence diagram evaluation—HUGIN, Shafer-Shenoy, and Lazy Propagation—highlighting how Lazy Evaluation significantly reduces computational complexity on LIMID diagrams.

Contribution

The paper demonstrates the advantages of Lazy Evaluation over traditional methods in influence diagram evaluation, especially on LIMID diagrams with explicit requisite information.

Findings

Lazy Evaluation offers significant computational savings.

Evaluation on LIMID diagrams benefits from explicit requisite information.

Lazy Propagation outperforms other architectures in efficiency.

Abstract

In this paper we compare three different architectures for the evaluation of influence diagrams: HUGIN, Shafer-Shenoy, and Lazy Evaluation architecture. The computational complexity of the architectures are compared on the LImited Memory Influence Diagram (LIMID): a diagram where only the requiste information for the computation of the optimal policies are depicted. Because the requsite information is explicitly represented in the LIMID the evaluation can take advantage of it, and significant savings in computational can be obtained. In this paper we show how the obtained savings is considerably increased when the computations performed on the LIMID is according to the Lazy Evaluation scheme.