Pareto Optimization or Cascaded Weighted Sum: A Comparison of Concepts

TL;DR

This paper compares Pareto optimization and cascaded weighted sum methods for multi-objective nonlinear problems, analyzing their suitability, strengths, and weaknesses in various real-world application scenarios.

Contribution

It introduces a classification of application scenarios and evaluates the applicability of Pareto and cascaded weighted sum methods across these scenarios.

Findings

Pareto optimization is dominant for small objectives but less suitable for many-objective problems.

Cascaded weighted sum offers advantages in automated and repetitive optimization tasks.

The paper discusses the strengths and weaknesses of both methods in different contexts.

Abstract

According to the published papers and books since the turn of the century, Pareto optimization is the dominating assessment method for multi-objective nonlinear optimization problems treated by population-based optimizers like Evolutionary Algorithms. However, is it always the method of choice for real-world applications, where either more than four objectives have to be considered, or the same type of task is repeated again and again with only minor modifications, in an automated optimization or planning process? This paper presents a classification of application scenarios and compares the Pareto approach with an extended version of the weighted sum, called cascaded weighted sum, for the different scenarios. Its range of application within the field of multi-objective optimization is discussed as well as its strengths and weaknesses.