Manage Events
All events are in Central time unless specified.
Colloquium

Mathematics Colloquium

Date:
Time:
4:00 pm – 4:50 pm
Avery Hall Room: 115
1144 T St
 Lincoln NE 68508
Additional Info: AVH
Contact:
Steve Cohn, (402) 472-7223, scohn1@math.unl.edu
Speaker: Taro Toyoizumi
Affiliation: RIKEN http://toyoizumilab.brain.riken.jp
Local Host: Vladimir Itskov
Title: Signal processing in neural networks that generate or receive noise.

Additional Public Info:

Abstract: Cortical neurons exhibit irregular activity patterns. However, the source of the cortical variability is unknown. Here, we study two different types of randomly connected networks of quadratic integrate-and-fire neurons that produce irregular spontaneous activity patterns: (a) a network model that has strong synaptic interactions and actively generates variability by chaotic nonlinear dynamics and (b) a network model that has weak synaptic interactions and receives noisy input, for example, by stochastic vesicle releases. These networks of spiking neurons are analytically tractable in the limit of a large network-size and channel-time-constant for integrating input. Despite the difference in their sources of variability, these two models can behave almost identically in their baseline activity. Indeed, when parameters are set appropriately, the two models are indistinguishable based on their spontaneous activity patterns unless majority of neurons in the network are simultaneously recorded. In spite of the close similarity in their spontaneous activity patterns, the two models can exhibit remarkably different sensitivity to external input. External input to the former network can reverberate within the network and be successfully read out over long time because of the strong synaptic interactions between neurons. On the other hand, input to the latter network rapidly decays because of the small synaptic interactions. The difference between the two types of network is further enhanced if synapses that provide external input to the networks undergo activity-dependent plasticity, producing a marked difference in the ability to decode external input from neural activity. We describe implications of this difference on population coding and neural plasticity. (Joint work with Isao Nishikawa, Kazuyuki Aihara).

Refreshments will be served in 348 Avery 3:30-4:00. The talk is free and open to the public.

Download this event to my calendar

This event originated in Math Colloquia.