ORCID

Abstract

The comprehension of the Evolutionary Algorithm (EA) search process is often eluded by challenges of transparency inherent to \textit{black-box} EAs, thus affecting algorithm enhancement and hyper-parameter optimisation. In this work, we develop algorithm insight by introducing the Population Dynamics Plot (PopDP). PopDP is a novel and intuitive visualisation capable of visualising the population of solutions, the parent-offspring lineage, solution perturbation operators, and the search process journey. We apply PopDP to NSGA-II to demonstrate the insight attained and the effectiveness of PopDP for visualising algorithm search on a series of discrete dual- and many-objective knapsack problems of different complexities, and our results demonstrate that the method can be used to produce a visualisation in which the lineage of solutions can be clearly seen. We also consider the efficacy of the proposed explainable visualisation against emerging approaches to benchmarking explainable AI methods and consider the accessibility of the resulting visualisations.

Publication Date

2022-07-19

Event

ECXAI — Evolutionary Computation and Explainable AI Workshop of GECCO 2022

Publication Title

GECCO'22: Proceedings of the Genetic and Evolutionary Computation Conference Companion

Acceptance Date

2022-04-25

Deposit Date

2022-03-05

Embargo Period

2022-07-22

Keywords

Evolutionary Computation, Explainability, Many-Objective Optimisation, Multi-Objective Optimisation, Visualisation

First Page

1794

Last Page

1802

Download

Share

COinS