Information-Theoretic Analysis for Understanding the Behavior of Song Learning by the Bengalese Finch

概要

論文の詳細を見る
Songbirds have been actively studied for their complex brain mechanism of sensor-motor integration during song learning. Male Bengalese finches learn singing by imitating external models. Birdsongs are strings of sounds represented by a sequence of letters called song notes. In this study, we focused on information-theoretic analysis of these sequential data to explore the complexity and diversity of birdsong throughout song development. We present ethological data mining results for birdsong, which showed that there are 2 types of song development: linear and non-linear. We found that all male birds in the early stages of development sing complex songs, which are gradually crystallized by the elimination of extra transitions. In contrast, this is not true in the case of non-linear song development, wherein there are no practice modes, and all activity counts toward selecting, constructing, and maintaining behavioral outcomes. We also found that contour surface diagrams of the transition matrix can be a good visual representation to distinguish song features. Our results indicate that information-theoretic analysis of behavioral sequences is important for the discovery of new aspects related to behavioral science.
2010-12-09