Multi-Objective AI Planning: Comparing Aggregation and Pareto Approaches

TL;DR

This paper compares aggregation-based and Pareto-based multi-objective approaches in AI planning, demonstrating the advantages of Pareto methods through experiments on a new benchmark set.

Contribution

It provides a comparative analysis of Pareto and aggregation approaches in multi-objective AI planning, validating the effectiveness of Pareto methods.

Findings

Pareto-based approaches outperform aggregation methods in certain benchmarks.

Divide and Evolve can be effectively adapted for multi-objective optimization.

Experimental results support the use of Pareto methods in multi-objective AI planning.

Abstract

Most real-world Planning problems are multi-objective, trying to minimize both the makespan of the solution plan, and some cost of the actions involved in the plan. But most, if not all existing approaches are based on single-objective planners, and use an aggregation of the objectives to remain in the single-objective context. Divide and Evolve (DaE) is an evolutionary planner that won the temporal deterministic satisficing track at the last International Planning Competitions (IPC). Like all Evolutionary Algorithms (EA), it can easily be turned into a Pareto-based Multi-Objective EA. It is however important to validate the resulting algorithm by comparing it with the aggregation approach: this is the goal of this paper. The comparative experiments on a recently proposed benchmark set that are reported here demonstrate the usefulness of going Pareto-based in AI Planning.