The inverse problem for neuronal networks is to infer its topology from analyzing its dynamics. Recently, transfer entropy[1], an information theoretical measure of directed interactions has become more popular for solving the inverse problem[2]. Due to its model-free nature, it can easily be applied to data in a variety of fields such as neuroscience, physiology, climate research and financial markets.

As observed in my experiments, "achromatic" receptive fields may react similarly to blue or green light stimuli, which are the two observable colors in bullfrogs. However, under intermeshed blue and green light stimulus, we constantly (such experiment had been repeated for more than five times) see the activity of recorded ganglion cells synchronize to only one color. I suspect that the retina randomly "follow" a color, since the same retina could produce activity following blue or green light under a same stimulus, and uninfluenced by the starting color.

Linear-nonlinear(LN) model is a approximation method for linking the time trace of stimulus with the firing rate. However, this is only a approximation and we can actually have some intuitive understanding about when it will fail.

Group meeting brief: July 17, 2014

CKC gave an overview of the objectives of our neural projects:

• How to grow a neuronal network (circuit)
• Nonlinear Properties of (small scale) networks
• Relation between structures and functions of networks (dynamics)
• Mechanism of Short term Memory (data read/write)
• Coding and processing of information in retina (psycho-physical → physiological → biophysical)
• Neuroprosthetic retina

* Please login to see his talk slides below.

Group meeting brief: June 24, 2014