2015-16: CCNS: Mathematics of Neural Networks and Neural Codes: March 9-11, 2016


Participation is by invitation only.
This workshop will be held at SAMSI in Research Triangle Park, NC.


This workshop will feature a full day of talks, targeting specific areas in the theory of neural networks and neural codes that are currently active areas of research. The second and third days will be focused on research sessions to further collaboration among the participants.

On the networks side, the focus is on training and fitting various types of networks (discrete models, rate models, spiking models) to produce dynamic activity patterns. This necessitates both a deep understanding of the dynamics of the networks in question, as well as state-of-the-art fitting techniques. Recent theoretical advances have led to an improved understanding of the dynamics of threshold-linear network, spiking networks with embedded cell assemblies, and pattern storage in Hopfield networks. Data sets for training these networks consist of synthetic spiking data, generated by spiking networks where we can control the “ground truth,” and real spiking data from electrophysiological recordings.

On the neural coding side, the emphasis is on the further development of topological and algebraic methods for analyzing neural activity from simultaneous recordings of neurons. This activity can be pre-processed to obtain neural correlations and neural codes. Many mathematical questions arise when trying to understand how important features of correlation or coding structure can be extracted from these data. The workshop will explain recent advances to participants, and build on existing research collaborations.

Local researchers from Duke, UNC, and NC State are welcome to attend the presentations.

Questions: email CCNS@samsi.info

Schedule and Supporting Media


Wednesday, March 9, 2016

Time Description Speaker Slides Videos
9:00-9:15 a.m. Registration
9:15-9:30 Opening Remarks
Morning Session – Recurrent Neural Networks
9:30-10:00 Overview of Network Models (Discrete, Rate, Spiking) Chris Hillar, Univ. of California, SF and Berkeley
10:00-11:00 Dynamics of Threshold-Linear Networks Carina Curto, Penn State University
11:00-12:00 Dynamics of Spiking Networks Brent Doiron, University of Pittsburgh
12:00 Fitting Recurrent Networks Chris Hillar, Univ. of California, SF and Berkeley
1:00-2:30 Lunch
Afternoon session – Algebra and Topology of Neural Codes
2:30-3:00 Overview of Topology in Neuroscience Carina Curto, Penn State University
3:00-4:00 Topological Analysis of Neural Correlations Chad Guisti, University of Pennsylvania
4:00-5:00 Convex Neural Codes Vladimir Itskov, Penn State University
5:00-6:00 Algebraic Analysis of Neural Codes Using the Neural Ring Nora Youngs, Harvey Mudd College

Thursday, March 10, 2016

Time Description Speaker Slides Videos
9:30-10:30 a.m. Discussion of Network-related Research Projects Led by Carina Curto, Brent Doiron and Chris Hillar
10:30-1:00 Collaborative Research Time
1:00-2:00 Lunch
2:00-3:00 Discussion of Topology-Related Research Projected Led by Chad Guisto and Vladimir Itskov
3:00-5:30 Collaborative Research Time

Friday, March 11, 2016

Time Description Speaker Slides Videos
9:30-12:00 p.m. Research Sessions
12:00-1:00 Lunch
1:00-3:00 Collaborative Research Time
3:00 Workshop Adjourns