Unsupervised learning to quantify differences in song learning of experimental zebra finch populations

Zebra finch song learning is a common model of motor learning processes, but quantification of song properties is lacking, particularly for comparison of experimental populations across development. Sparse convolutional feature extraction, a method previously used to analyze other natural sounds, is applied to zebra finch song here. The results of feature extraction were used to develop metrics that were applied to zebra finch song from across both normal and isolate development. As expected, adult control song was substantially different from adult isolate song in all metrics. More interestingly, differences in some metrics were seen between the two as early in development as recordings were taken, suggesting that differences exist prior to obvious abnormalities appearing in the song spectrogram. Overall, these results provide interesting ideas about isolate song learning, and act as a proof of concept for the use of sparse convolutional learning to compare bird populations.

Department dc:contributor.department

Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science.

Grantor dc:publisher

Massachusetts Institute of Technology

Year dc:date.issued

2017

Author dc:creator

• Ennis, Michaela (Michaela M.)

Advisor dc:contributor.advisor

• Michale S. Fee.

• Electrical Engineering and Computer Science.

Statement dc:rights

• MIT theses are protected by copyright. They may be viewed, downloaded, or printed from this source but further reproduction or distribution in any format is prohibited without written permission.

Licence dc:rights.uri

• http://dspace.mit.edu/handle/1721.1/7582

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/1721.1/119521

OAI identifier oai:identifier

oai:dspace.mit.edu:1721.1/119521