Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Relevance Score: A Landmark-Like Heuristic for Planning

Oliver Kim, Mohan Sridharan

TL;DR

While the original landmark-based heuristic leads to better performance on problems with well-defined landmarks, this approach substantially improves performance on problems that lack non-trivial landmarks.

Abstract

Landmarks are facts or actions that appear in all valid solutions of a planning problem. They have been used successfully to calculate heuristics that guide the search for a plan. We investigate an extension to this concept by defining a novel "relevance score" that helps identify facts or actions that appear in most but not all plans to achieve any given goal. We describe an approach to compute this relevance score and use it as a heuristic in the search for a plan. We experimentally compare the performance of our approach with that of a state of the art landmark-based heuristic planning approach using benchmark planning problems. While the original landmark-based heuristic leads to better performance on problems with well-defined landmarks, our approach substantially improves performance on problems that lack non-trivial landmarks.

Relevance Score: A Landmark-Like Heuristic for Planning

TL;DR

While the original landmark-based heuristic leads to better performance on problems with well-defined landmarks, this approach substantially improves performance on problems that lack non-trivial landmarks.

Abstract

Landmarks are facts or actions that appear in all valid solutions of a planning problem. They have been used successfully to calculate heuristics that guide the search for a plan. We investigate an extension to this concept by defining a novel "relevance score" that helps identify facts or actions that appear in most but not all plans to achieve any given goal. We describe an approach to compute this relevance score and use it as a heuristic in the search for a plan. We experimentally compare the performance of our approach with that of a state of the art landmark-based heuristic planning approach using benchmark planning problems. While the original landmark-based heuristic leads to better performance on problems with well-defined landmarks, our approach substantially improves performance on problems that lack non-trivial landmarks.
Paper Structure (21 sections, 21 equations, 3 figures, 2 tables, 2 algorithms)

This paper contains 21 sections, 21 equations, 3 figures, 2 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related work
  3. Definitions and Background Knowledge
  4. Classical planning problem
  5. Landmarks
  6. Simplifying approximations
  7. Backtracking tree
  8. Non-deterministic agent
  9. Relevance score
  10. Choices counter
  11. Lowest common ancestors
  12. Calculating the Relevance Score
  13. Derivation
  14. Exploration
  15. Finding LCAs
  16. ...and 6 more sections

Figures (3)

  • Figure 1: Illustration of the distinction between landmarks and relevant factsp1 is a landmark because all partial plans that achieve the goal g must use one of actions A1, A2, or A3, for which it is a precondition. p2 is not a landmark because the goal could be achieved by A1, but it is highly relevant because there is a $\frac{2}{3}$ chance of it being present in any partial plan to achieve the goal.
  • Figure 2: Standard problems (674): for each measure, a lower value is better. S2 (landmark-counting) is better than S1 (relevance score) based on most measures; results support hypothesis H1.
  • Figure 3: Landmark free problems (500): for each measure, a lower value is better. S1 (relevance score) is better than S2 (landmark-counting) based on most measures; results support hypothesis H2.