Presentation Details
| Machine learning animal behaviors from field and crowd-sourced imagery Andrew Hein. Cornell University, Department of Computational Biology, Ithaca, NY, USA |
Abstract
Fields like systems neuroscience and computational ethology are increasingly relying on high-dimensional, precision quantification of animal behavior in laboratory experiments to understand the mechanisms and logic of animal behavior. These measurements leverage advancements in machine learning that seek to discover structure in behavioral sequences de novo from data. I will discuss how similar strategies can be used to study behavior of wild animals using imagery from the field and from open online nature imagery platforms. Using collective behavior and predator-prey interactions as case studies, I will show how these approaches can provide fundamental new insights about the behaviors that shape fitness in the wild.
No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.
No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.