Data Visualization Scenarios for the Analysis of Computational Evolutionary Techniques

There has been an increasing demand to understand and describe Evolutionary Computing techniques. Information Visualization may contribute with interactive data visualizations that help explore the population of individual solutions over the data search space and generations, convergent behavior, in...

Full description

Saved in:
Bibliographic Details
Main Authors: Santa Rosa Nassar Dos Santos, Yuri, Simoes Meiguins, Aruanda, Hortencio Dos Santos, Diego, Gustavo Resque Dos Santos, Carlos, Magalhaes De Morais, Jefferson, Serique Meiguins, Bianchi
Format: Conference Proceeding
Language:English
Subjects:
Clustering algorithms
Data visualization
Evolutionary Algorithms
Evolutionary computation
Genetic algorithms
Information Visualization
Sociology
Statistics
Visualization Techniques
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:There has been an increasing demand to understand and describe Evolutionary Computing techniques. Information Visualization may contribute with interactive data visualizations that help explore the population of individual solutions over the data search space and generations, convergent behavior, individual fitness, the dynamic of the evolutionary process among other possible scenarios. Although there are previous works on the use of visualization to analyze evolutionary techniques, there has been little diversity among the approached visualization techniques. Also, most related works consider only genetic algorithms and ignore other evolutionary approaches. Therefore the goal of this paper is to suggest the appropriate InfoVis techniques for the analyzed scenarios to better understand the behavior of evolutionary computing algorithms. Furthermore, we present a case study that applies the proposed scenarios to AutoClustering, a tool based on Estimation of Distribution Algorithms. We hope the proposed scenarios and techniques provide a set of good practices for the analysis of Evolutionary Computing techniques.
ISSN:2375-0138
DOI:10.1109/IV.2019.00056