Distance analysis of music melodies created by distributed interactive GA

In the area of product design, adding user's kansei (feelings) on the products is important for emphasizing its value. Interactive Evolutionary Computation (IEC) is well known method that searches optimal or better media contents suited for each user's kansei. Some previous studies applied...

Full description

Saved in:
Bibliographic Details
Published in2017 International Conference on Biometrics and Kansei Engineering (ICBAKE) pp. 80 - 83
Main Authors Nomura, Kota, Fukumoto, Makoto
Format Conference Proceeding
LanguageEnglish
Japanese
Published IEEE 01.09.2017
Subjects
Computational modeling
Distance
Distributed Interactive Genetic Algorithm
Evolutionary computation
Genetic algorithms
IEC
Interactive Genetic Algorithm
Multiple users
Music
Music melody
Sociology
Statistics
Online AccessGet full text
DOI10.1109/ICBAKE.2017.8090621

Cover

More Information
Summary:In the area of product design, adding user's kansei (feelings) on the products is important for emphasizing its value. Interactive Evolutionary Computation (IEC) is well known method that searches optimal or better media contents suited for each user's kansei. Some previous studies applied IEC into problem of multiple users. Purpose of the present study is to compare Distributed Interactive Genetic Algorithm (DIGA) and Interactive Genetic Algorithm (IGA) in a view point of creating music melody. We conducted listening experiments for investigating efficiency of the DIGA and IGA for creation of music melody. Ten persons were divided into five pairs, the pair participated in the experiment as subjects simultaneously. As an index of searching space of the pair, distance of melodies between the subjects in the pair in each generation was investigated. As a result, a trend of decrease in the distance was observed in the DIGA where exchange of the best individuals between the subjects.
DOI:10.1109/ICBAKE.2017.8090621