A new data-mining method to search for behavioral properties that induce alignment and their involvement in social learning in medaka fish (Oryzias latipes).

Takashi Ochiai, Yuji Suehiro, Katsuhiro Nishinari, Takeo Kubo, Hideaki Takeuchi

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

Coordinated movement in social animal groups via social learning facilitates foraging activity. Few studies have examined the behavioral cause-and-effect between group members that mediates this social learning. We first established a behavioral paradigm for visual food learning using medaka fish and demonstrated that a single fish can learn to associate a visual cue with a food reward. Grouped medaka fish (6 fish) learn to respond to the visual cue more rapidly than a single fish, indicating that medaka fish undergo social learning. We then established a data-mining method based on Kullback-Leibler divergence (KLD) to search for candidate behaviors that induce alignment and found that high-speed movement of a focal fish tended to induce alignment of the other members locally and transiently under free-swimming conditions without presentation of a visual cue. The high-speed movement of the informed and trained fish during visual cue presentation appeared to facilitate the alignment of naïve fish in response to some visual cues, thereby mediating social learning. Compared with naïve fish, the informed fish had a higher tendency to induce alignment of other naïve fish under free-swimming conditions without visual cue presentation, suggesting the involvement of individual recognition in social learning. Behavioral cause-and-effect studies of the high-speed movement between fish group members will contribute to our understanding of the dynamics of social behaviors. The data-mining method used in the present study is a powerful method to search for candidates factors associated with inter-individual interactions using a dataset for time-series coordinate data of individuals.

Original languageEnglish
Article numbere71685
JournalPLoS One
Volume8
Issue number9
Publication statusPublished - 2013
Externally publishedYes

Fingerprint

Oryzias
Oryzias latipes
Data Mining
Fish
Data mining
Fishes
learning
fish
Cues
methodology
Social Learning
Food
Social Behavior
social behavior
Reward

ASJC Scopus subject areas

  • Agricultural and Biological Sciences(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Medicine(all)

Cite this

A new data-mining method to search for behavioral properties that induce alignment and their involvement in social learning in medaka fish (Oryzias latipes). / Ochiai, Takashi; Suehiro, Yuji; Nishinari, Katsuhiro; Kubo, Takeo; Takeuchi, Hideaki.

In: PLoS One, Vol. 8, No. 9, e71685, 2013.

Research output: Contribution to journalArticle

@article{56eb316d81cb47d08de7b9d3ea625118,
title = "A new data-mining method to search for behavioral properties that induce alignment and their involvement in social learning in medaka fish (Oryzias latipes).",
abstract = "Coordinated movement in social animal groups via social learning facilitates foraging activity. Few studies have examined the behavioral cause-and-effect between group members that mediates this social learning. We first established a behavioral paradigm for visual food learning using medaka fish and demonstrated that a single fish can learn to associate a visual cue with a food reward. Grouped medaka fish (6 fish) learn to respond to the visual cue more rapidly than a single fish, indicating that medaka fish undergo social learning. We then established a data-mining method based on Kullback-Leibler divergence (KLD) to search for candidate behaviors that induce alignment and found that high-speed movement of a focal fish tended to induce alignment of the other members locally and transiently under free-swimming conditions without presentation of a visual cue. The high-speed movement of the informed and trained fish during visual cue presentation appeared to facilitate the alignment of na{\"i}ve fish in response to some visual cues, thereby mediating social learning. Compared with na{\"i}ve fish, the informed fish had a higher tendency to induce alignment of other na{\"i}ve fish under free-swimming conditions without visual cue presentation, suggesting the involvement of individual recognition in social learning. Behavioral cause-and-effect studies of the high-speed movement between fish group members will contribute to our understanding of the dynamics of social behaviors. The data-mining method used in the present study is a powerful method to search for candidates factors associated with inter-individual interactions using a dataset for time-series coordinate data of individuals.",
author = "Takashi Ochiai and Yuji Suehiro and Katsuhiro Nishinari and Takeo Kubo and Hideaki Takeuchi",
year = "2013",
language = "English",
volume = "8",
journal = "PLoS One",
issn = "1932-6203",
publisher = "Public Library of Science",
number = "9",

}

TY - JOUR

T1 - A new data-mining method to search for behavioral properties that induce alignment and their involvement in social learning in medaka fish (Oryzias latipes).

AU - Ochiai, Takashi

AU - Suehiro, Yuji

AU - Nishinari, Katsuhiro

AU - Kubo, Takeo

AU - Takeuchi, Hideaki

PY - 2013

Y1 - 2013

N2 - Coordinated movement in social animal groups via social learning facilitates foraging activity. Few studies have examined the behavioral cause-and-effect between group members that mediates this social learning. We first established a behavioral paradigm for visual food learning using medaka fish and demonstrated that a single fish can learn to associate a visual cue with a food reward. Grouped medaka fish (6 fish) learn to respond to the visual cue more rapidly than a single fish, indicating that medaka fish undergo social learning. We then established a data-mining method based on Kullback-Leibler divergence (KLD) to search for candidate behaviors that induce alignment and found that high-speed movement of a focal fish tended to induce alignment of the other members locally and transiently under free-swimming conditions without presentation of a visual cue. The high-speed movement of the informed and trained fish during visual cue presentation appeared to facilitate the alignment of naïve fish in response to some visual cues, thereby mediating social learning. Compared with naïve fish, the informed fish had a higher tendency to induce alignment of other naïve fish under free-swimming conditions without visual cue presentation, suggesting the involvement of individual recognition in social learning. Behavioral cause-and-effect studies of the high-speed movement between fish group members will contribute to our understanding of the dynamics of social behaviors. The data-mining method used in the present study is a powerful method to search for candidates factors associated with inter-individual interactions using a dataset for time-series coordinate data of individuals.

AB - Coordinated movement in social animal groups via social learning facilitates foraging activity. Few studies have examined the behavioral cause-and-effect between group members that mediates this social learning. We first established a behavioral paradigm for visual food learning using medaka fish and demonstrated that a single fish can learn to associate a visual cue with a food reward. Grouped medaka fish (6 fish) learn to respond to the visual cue more rapidly than a single fish, indicating that medaka fish undergo social learning. We then established a data-mining method based on Kullback-Leibler divergence (KLD) to search for candidate behaviors that induce alignment and found that high-speed movement of a focal fish tended to induce alignment of the other members locally and transiently under free-swimming conditions without presentation of a visual cue. The high-speed movement of the informed and trained fish during visual cue presentation appeared to facilitate the alignment of naïve fish in response to some visual cues, thereby mediating social learning. Compared with naïve fish, the informed fish had a higher tendency to induce alignment of other naïve fish under free-swimming conditions without visual cue presentation, suggesting the involvement of individual recognition in social learning. Behavioral cause-and-effect studies of the high-speed movement between fish group members will contribute to our understanding of the dynamics of social behaviors. The data-mining method used in the present study is a powerful method to search for candidates factors associated with inter-individual interactions using a dataset for time-series coordinate data of individuals.

UR - http://www.scopus.com/inward/record.url?scp=84902040251&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84902040251&partnerID=8YFLogxK

M3 - Article

C2 - 24039720

AN - SCOPUS:84902040251

VL - 8

JO - PLoS One

JF - PLoS One

SN - 1932-6203

IS - 9

M1 - e71685

ER -