Georgia Institute of Technology College of Sciences

This is a part of the GT MAP activities on Control.  GT MAP is a place for research discussion and collaboration. We welcome participation of any researcher interested in discussing his/her project and exchange ideas with Mathematicians.

There will be light refreshments through out the event. This seminar will be held in Skiles 006 and refreshments at Skiles Atrium.

A couple of members of Prof. Pandarinath's group will present their research

3:00 PM - 3:45PM Prof.  Pandarinath will give a talk on "Unsupervised discovery of ensemble dynamics in the brain using deep learning techniques."

3:45PM -- 4:00PM Break with Discussions

4:00PM - 4:25PM Second talk given by Lahiru Wimalasena (Graduate student, Wallace H. Coulter Department of Biomedical Engineering, Emory University and Georgia Tech).

4:25PM - 5PM Discussion of open problems stemming from the presentations.

Title:  Unsupervised discovery of ensemble dynamics in the brain using deep learning techniques

Abstract: 

Since its inception, neuroscience has largely focused on the neuron as the functional unit of the nervous system. However, recent evidence demonstrates that populations of neurons within a brain area collectively show emergent functional properties ("dynamics"), properties that are not apparent at the level of individual neurons. These emergent dynamics likely serve as the brain’s fundamental computational mechanism. This shift compels neuroscientists to characterize emergent properties – that is, interactions between neurons – to understand computation in brain networks. Yet this introduces a daunting challenge – with millions of neurons in any given brain area, characterizing interactions within an area, and further, between brain areas, rapidly becomes intractable.

I will demonstrate a novel unsupervised tool, Latent Factor Analysis via Dynamical Systems ("LFADS"), that can accurately and succinctly capture the emergent dynamics of large neural populations from limited sampling. LFADS is based around deep learning architectures (variational sequential auto-encoders), and builds a model of an observed neural population's dynamics using a nonlinear dynamical system (a recurrent neural network).

When applied to neuronal ensemble recordings (~200 neurons) from macaque primary motor cortex (M1), we find that modeling population dynamics yields accurate estimates of the state of M1, as well as accurate predictions of the animal's motor behavior, on millisecond timescales. I will also demonstrate how our approach allows us to infer perturbations to the dynamical system (i.e., unobserved inputs to the neural population), and further allows us to leverage population recordings across long timescales (months) to build more accurate models of M1's dynamics.

This approach demonstrates the power of deep learning tools to model nonlinear dynamical systems and infer accurate estimates of the states of large biological networks. In addition, we will discuss future directions, where we aim to pry open the "black box" of the trained recurrent neural networks, in order to understand the computations being performed by the modeled neural populations.

pre-print available: lfads.github.io

This is a part of the GT MAP activities on Control.  GT MAP is a place for research discussion and collaboration. We welcome participation of any researcher interested in discussing his/her project and exchange ideas with Mathematicians.

There will be light refreshments through out the event. This seminar will be held in Skiles 006 and refreshments at Skiles Atrium.

A couple of members of Prof. Theodorou's group will present their research

3:00 PM - 3:45PM Prof. Theodorou will give a talk on " The science of autonomy: A "happy" symbiosis between learning, control and physics."

3:45PM -- 4:00PM Break with Discussions

4:00PM - 4:25PM another talk.

4:25PM - 5PM Discussion of open problems stemming from the presentations.

Title: The science of autonomy: A "happy" symbiosis between learning, control and physics.

Abstract:   In this talk I will present an information theoretic approach to stochastic optimal control and inference  that has advantages  over classical methodologies and theories for decision making under uncertainty.  The main idea  is that there are certain connections between optimality principles in control and information theoretic inequalities in statistical physics that allow  us to solve hard decision making problems in robotics, autonomous systems and beyond. There are essentially two different points of view  of the same "thing" and these two different points of view  overlap   for a fairly general class of dynamical systems that undergo stochastic effects.  I will also present a holistic view of autonomy that collapses planning, perception and control into one computational engine, and ask questions  such as  how organization and structure relates to computation and performance. The last part of my talk includes computational frameworks for uncertainty representation   and suggests ways to incorporate these representations within decision making and control.

Four minicourses of one hour each will be given by Professor Marek Biskup of UCLA and Professor Jonathan Mattingly of Duke.

Invited Speakers

• Louis-Pierre Arguin (CUNY)
• Rodrigo Bañuelos (Purdue)
• Fabrice Baudoin (Connecticut)
• Gerandy Brito (Georgia Tech)
• Shirshendu Ganguly (Berkeley)
• Jack Hanson (CUNY)
• David Herzog (Iowa State)
• Jessica Lin (McGill)
• Firas Rassoul-Agha (Utah)
• Philippe Sosoe (Cornell)

Topics

• Spatial stochastic systems 
• Percolation and spin systems 
• Growth models
• Random matrices 
• Stochastic differential equations
• Stochastic homogenization.

Sponsors

• National Science Foundation
• Institute for Mathematics and Its Applications

Local Organizers

• Christian Houdré
• Michael Damron

More information at the conference website, https://pwp.gatech.edu/rtip/

Register by Mar. 1, 2018

If you would like to give a talk or obtain support for lodging, kindly submit your application by Mar. 1, 2018.  Please email the organizers Christian Houdré and Michael Damron.

School of Math welcomes admitted PhD students as they visit campus to get to know the faculty, and interact with current students.

CRIDC aims to equip Georgia Tech graduate students with tools and insights to survive and thrive in the ever-changing technological, financial, and political landscapes of the 21st-century employment market. The conference will provide graduate students of all majors and degree types opportunities to explore varied career paths.

The 2018 CRIDC features a new component – A Graduate Career Fair,  which is tailored to the needs of Georgia Tech graduate students.

2018 CRIDC features the following events:

• Panels (differentiated into industry, consulting, academia, entrepreneurship, and government/national labs)
• Company Talks (informational sessions of various companies)
• Workshops
• Poster Competition
• VentureLab Innovation Competition
• Graduate Career Fair (new!)
• Additional Networking Opportunities

Visit the CRIDC Website for more information.

EDITOR'S NOTE: This event was first announced on the Georgia Tech campus calendar. For updates, check the original posting. 

Flashpoint works closely with founders to enable them to think clearly about their businesses. It is unique in implementing startup engineering, a business creation and innovation process developed by Merrick Furst, Distinguished Professor in the College of Computing at Georgia Tech.

For the second year in a row, Flashpoint is hosting the Innovation for All Conference. The theme for 2018 is Building a Deliberately Innovative Culture. Participants will learn how entrepreneurs, large enterprises, and educational institutions use deliberate innovation practices to avoid common failure paths and innovate reliably.

The conference will begin with a panel discussion moderated by Rich A. DeMillo, director of the Center for 21st Century Universities  at Georgia Tech.

Among the panel discussants are two faculty members from the College of Sciences: Lew Lefton and Michael Schatz. Lefton is Georgia Tech assistant vice president for research cyberinfrastructure, College of Sciences assistant dean for information technology, and School of Mathematics senior academic professional. Schatz is a professor in the School of Physics.

The conference includes a demonstration of Flashpoint techniques in a master class-type setting, with innovation teams from startups, large companies and from Georgia Tech. 

Register at http://flashpoint.co/demo-days/

The 2018 Meeting on Applied Algebraic Geometry (MAAG 2018) will be hosted by the School of Mathematics at Georgia Institute of Technology on April 7-8, 2018. 

MAAG is a regional gathering that attracts participants primarily from the Southeast of the United States. Previous meetings took place at Georgia Tech in 2015 and at Clemson University in 2016.

Organizers of MAAG 2018 have invited several speakers from outside the Southeastern U.S. and are open to "longer distance" participants as well. Some funding is available to participate (see registration form, priority is given to students).

The main event on Saturday, April 7, will be followed by an informal Numerical Algebraic Geometry day on Sunday, April 8, which all participants are encouraged to attend.

Registration is free until March 1. After March 1, 2018, registration is $40.

Register here: https://sites.google.com/view/maag-2018

Speakers on April 7 

• Dan Bates (Colorado State)
• Florian Enescu (Georgia State)
• Jon Hauenstein (Notre Dame)
• Kaie Kubjas (MIT/Aalto)
• Vicki Powers (Emory)
• Seth Sullivant (North Carolina State)

Organizers on April 8

• Dan Bates (Colorado State)
• Jon Hauenstein (Notre Dame)
• Anton Leykin (Georgia Tech)

Algebraic geometers in Georgia and neighboring southern states gather at the Georgia Tech School of Mathematics on Feb. 23-25, 2018, to strengthen their growing community. The goal is to support early-career mathematicians, especially those from groups that are underrepresented in mathematics. The gathering – the 2018 Georgia Algebraic Geometry Symposium (GAGS) – will provide attendees opportunities to network and be exposed to cutting-edge developments in the field.

Classical algebraic geometry studies questions about solutions sets of polynomial equations. For example, what is the dimension of the solution space? What is its shape? Can we break up the solution space into simpler components? Modern algebraic geometry includes much more abstract objects. 

Algebraic geometry has connections to complex analysis, string theory, topology, number theory, and game theory. It has applications in statistics, robotics, phylogenetics, and geometric modeling.

The Georgia Tech School of Mathematics is home to several experts in algebraic geometry: faculty members Matthew Baker, Greg Blekherman, Anton Leykin, Joseph Rabinoff, Kirsten Wickelgren, and Josephine Yu; postdoctoral fellows Padmavathi Srinivasan and Philipp Jell; and senior academic professional Salvador Barone.

Baker is organizing 2018 GAGS, with assistance from Rabinoff and Josephine Yu. The College of Sciences is providing additional financial support.

The 2018 symposium in Georgia Tech is the culmination of a three-year National Science Foundation (NSF) grant (DMS-1529573) to Georgia Tech, University of Georgia (UGA), and Emory University to organize and host GAGS in rotation over three years. The symposium was held at Emory University in 2016 and at UGA in 2017.

The 2018 GAGS features the following invited speakers, including one from Germany:

• Linda Chen (Swarthmore College)
• June Huh (Institute for Advanced Study)
• David Jensen (University of Kentucky)
• Jesse Kass (University of South Carolina)
• Lek-Heng Lim (University of Chicago)
• Kristin Shaw (Technischen Universität Berlin)
• Andrew Snowden (University of Michigan)
• Padmavathi Srinivasan (Georgia Institute of Technology)

Students are encouraged to attend. Registration is free.

Jan 16-19 2018

Edit: This workshop was also featured in this story from the College of Science.

Workshop on introduction to dynamical systems methods for mission design

This workshop is designed to provide an introduction to recent methods to design orbits for spacecraft which use less fuel by taking advantage of natural gravitational forces. The analysis of such missions requires a deep understanding of the structure of phase space.

The workshop is part of a SURP grant by JPL Graduate and undergraduate students are encouraged to attend.

Some information on the workshop is available at: http://people.math.gatech.edu/~rll6/JPL/jpl.html

Full article from the College of Science: http://www.news.gatech.edu/2018/01/12/mathematics-fuels-space-exploration

The Tech Topology Conference brings together established and beginning researchers from around the country for a weekend of mathematics in Atlanta. Check back soon for more details. We are pleased to announce this year's speakers:

• Tori Akin (Duke University)
• Lei Chen (University of Chicago)
• Diana Hubbard (University of Michigan)
• Tye Lidman (North Carolina State University)
• Ciprian Manolescu (UCLA)
• Emmy Murphy (Northwestern University)
• Balázs Strenner (Georgia Tech)

The 2017 conference features a session of five-minute lightning talks.

If you are interested in giving such a talk (on behalf of your work or someone else’s) please see the "Registration and Support" page.
Deadline for submitting proposals for Lightning Talks is October 31.

organizers: J. Etnyre, J. Hom, P. Lambert-Cole, J. Lanier, C. Leverson, D. Margalit, and B. Strenner
Supported by the NSF and the Georgia Institute of Technology