Python is for the birds and for the brains
Details
http://photos1.meetupstatic.com/photos/event/6/3/6/4/600_452845444.jpeg
David Nicholson, a graduate student in the neuroscience program at Emory University will speak on the application of python-based machine learning and analytics to the study of birdsong.
Doors will open at 7. Here is a Google Maps link to the Michael Street parking deck (https://www.google.com/maps/place/Michael+Street+Parking+Deck/@33.7969056,-84.3237508,18z/data=!4m5!3m4!1s0x88f506f7702a174f:0xfa8324343a79edf6!8m2!3d33.7965044!4d-84.3246949). The auditorium is on the ground floor of the Whitehead Biomedical Research Building (https://www.google.com/maps/place/Whitehead+Biomedical+Research+Building,+Atlanta,+GA+30329/@33.7958488,-84.325436,17z/data=!3m1!4b1!4m5!3m4!1s0x88f506f767eabafd:0x517ad4685ae50825!8m2!3d33.7958444!4d-84.3232473), right next to the parking.
Food and drinks will be provided by Emory Computational Neuroscience (http://biomed.emory.edu/PROGRAM_SITES/NS/research/computational-neuroscience.html). The main talk will start at 7:30, followed by a town-hall style Q&A session. We will then have several short lightning talks, including one about the application of machine learning to drug discovery (Tom Kaiser, Emory).
As always, please consider signing up for a lightning talk yourself: http://bit.ly/PyDataATL-lightning
Abstract:
Like many other fields, neuroscience faces an explosion of data and an array of questions about those data that can’t be answered with cookie cutter analyses. An increasing number of neuroscientists are turning to the programming language Python to develop new methods for analyzing data and for modeling the brain. Python provides many advantages to scientists. These include an easy-to-read syntax and a mature ecosystem of code and platforms for data analysis and modeling.
I will illustrate those advantages by discussing one of my projects in the Sober lab (http://www.biology.emory.edu/research/Sober/Home.html). We study songbirds as a model system for how the brain learns and produces complex skills, such as speech. Much like babies learning to talk from their parents, songbirds learn their song from a tutor during a critical period in development, and they then refine the song by practicing thousands of times.
Our lab investigates how the brain controls the muscles that produce this behavior, and how sensory feedback from behavior can in turn change the brain. We carry out experiments to study learning, each of which requires us to record song from an individual bird in our lab. We label the elements of its song, known as syllables, so we can analyze our results. A bird can sing hundreds of songs a day. To deal with this mountain of data, I have applied machine learning methods that automate labeling song syllables. I will discuss how Python has made it possible for me to conveniently compare the previous published methods for labeling syllables, and improve upon those methods.
To close I will briefly discuss the many other domains in neuroscience where Python is used. My goal will be to show why all (neuro)scientists need training in Python or similar languages like R. At the same time this whirlwind tour of brain studies should be of broad interest.
Sponsors
