Viterbi Faculty of Electrical Engineering, Technion
Optimal Encoding in Sensory Neural Populations
Populations of sensory neurons encode information about the external world through their spiking activity. To understand this encoding, it is natural to model it as an optimal or near-optimal code in the context of some task performed by higher brain regions, using performance criteria such as decoding error or motor performance. Decoding of sensory spike trains may be formulated as a nonlinear filtering problem based on point-process observations. A solution to this problem is useful to characterize optimal encoding.
The problem of optimal neural decoding has been solved analytically only for homogeneous neural populations that uniformly cover the stimulus space. Even in this simplified model, many previous works considered easier proxies to the encoding problem, such as maximization of Fisher information or minimization of decoding error for a non-optimal decoders.
We present an approximate solution to the decoding problem for more general sensory models. The approximate closed-form decoder provides insight into the nature of optimal neural codes that is not easily gleaned from numerical methods. We study optimal encoding both in the uniform case, where estimation error may be evaluated in closed form, and in the non-uniform case, where the approximate filter is used. Optimal codes are found to match experimentally observed neural codes. We show that when minimizing decoding error in a multivariate setting, additional constraints may be necessary to obtain a well-posed problem. Minimization under biologically plausible constraints yields solutions that qualitatively differ from those previously obtained by optimizing proxies to decoding error.
* PhD student supervised by Prof. Ron Meir.
Sun 17 Mar 2019
Start Time: 14:30
End Time: 15:30
430 | Fischbach Faculty Of Electrical Engineering