KNN-Averaging for Noisy Multi-objective Optimisation

  title={KNN-Averaging for Noisy Multi-objective Optimisation},
  author={Stefan Klikovits and Paolo Arcaini},
Multi-objective optimisation is a popular approach for finding solutions to complex problems with large search spaces that reliably yields good optimisation results. However, with the rise of cyber-physical systems, emerges a new challenge of noisy fitness functions, whose objective value for a given configuration is non-deterministic, producing varying results on each execution. This leads to an optimisation process that is based on stochastically sampled information, ultimately favouring… 

Handling Noise in Search-Based Scenario Generation for Autonomous Driving Systems

This paper uses kNN-Avg for the scenario generation of a real-world autonomous driving system (ADS) and shows that it is better than the noisy baseline and compares it to the repetition-method and outline indicators as to which approach to choose in which situations.



An Investigation on Noisy Environments in Evolutionary Multiobjective Optimization

  • C. GohK. Tan
  • Computer Science
    IEEE Transactions on Evolutionary Computation
  • 2006
Three noise-handling features are proposed based upon the analysis of empirical results, including an experiential learning directed perturbation operator that adapts the magnitude and direction of variation according to past experiences for fast convergence and a possibilistic archiving model based on the concept of possibility and necessity measures to deal with problem of uncertainties.

Evolutionary Multi-objective Ranking with Uncertainty and Noise

The ranking process needed to provide probabilities of selection is re-formulated to begin to account for the uncertainties and noise present in the system being optimised.

Genetic Algorithms in Noisy Environments

This work explores in detail the tradeoffs between the amount of effort spent on evaluating each structure and the number of structures evaluated during a given iteration of the genetic algorithm.

Pareto-Front Exploration with Uncertain Objectives

For objective values that are constrained by intervals, a theory of probabilistic dominance is derived, an extension of the definition of Pareto-dominance, and it is shown how this theory may be used in order to guide the selection process to approximate the Pare to-set.

Evolutionary Multi-objective Optimization in Uncertain Environments - Issues and Algorithms

  • C. GohK. Tan
  • Computer Science
    Studies in Computational Intelligence
  • 2009
This book is intended for a wide readership and will be a valuable reference for engineers, researchers, senior undergraduates and graduate students who are interested in the areas of evolutionary multi-objective optimization and uncertainties.

Pymoo: Multi-Objective Optimization in Python

This work develops pymoo, a multi-objective optimization framework in Python that addresses practical needs, such as the parallelization of function evaluations, methods to visualize low and high-dimensional spaces, and tools for multi-criteria decision making.

Quality Evaluation of Solution Sets in Multiobjective Optimisation

This article summarises and categorises 100 state-of-the-art quality indicators and discusses issues regarding attributes that indicators possess and properties that indicators are desirable to have, in the hope of motivating researchers and practitioners to look into these important issues when designing quality indicators.

Comparison of Multiobjective Evolutionary Algorithms: Empirical Results

This paper provides a systematic comparison of various evolutionary approaches to multiobjective optimization using six carefully chosen test functions and shows that elitism is shown to be an important factor for improving evolutionary multiobjectives search.

A comparative study of constrained multi-objective evolutionary algorithms on constrained multi-objective optimization problems

Among the eight CMOEAs, MOEA/D-IEpsilon with both SBX and DE operators has the best performance on the twenty-three test problems.

Creating Robust Solutions by Means of Evolutionary Algorithms

For real world problems it is often not sufficient to find solutions of high quality, but the solutions should also be robust. By robust we mean that the quality of the solution does not falter