Embedded Files
Oscillatory Neural Networks
For the first time ever, CNSLM is hosting a workshop!
For the first time ever, CNSLM is hosting a workshop!
This workshop will serve as an introduction to the field of inference-based generative modelling of neural dynamics. We will demonstrate the universal reconstruction ability of a hybrid network, constituting of a network of a primitive model of neural oscillator along with a feed-forward MLP framework residing on top of it.
About the Workshop
About the Workshop
A pair of coupled limit cycle harmonic oscillators can show a wide range of dynamical phenomena such as synchronization and entrainment depending on the coupling strategy. The first exercise is to analyze the transient and the steady state response of a pair of Hopf oscillators coupled through various kinds of coupling strategies qualitatively and quantitatively. Next we are going to build a generative model out of a network of such oscillators to reconstruct multidimensional EEG and fMRI signals. The universal reconstruction ability of such a simple reservoir model is investigated thereafter.
A pair of coupled limit cycle harmonic oscillators can show a wide range of dynamical phenomena such as synchronization and entrainment depending on the coupling strategy. The first exercise is to analyze the transient and the steady state response of a pair of Hopf oscillators coupled through various kinds of coupling strategies qualitatively and quantitatively. Next we are going to build a generative model out of a network of such oscillators to reconstruct multidimensional EEG and fMRI signals. The universal reconstruction ability of such a simple reservoir model is investigated thereafter.
Prerequisites
Prerequisites
Software: Matlab
Software: Matlab
Page updated
Google Sites
Report abuse