We are trying to figure out pairs of neurons are connected (through a directed edge) from the time series activity of the neurons.
- Understanding the topology underlying neural connections could unravel the cure of neuro-pathologies like epilepsy, Alzheimer's disease.
- It would also lead to a better understanding of how the barin functions in general.
- The task involved inferring statistical causality from time-series data, which is an exciting topic in itself.
- To see why the causality is interesting, go through the comic strips below. First two are from xkcd and the third (at the bottom) from SMBC.
We look at what different measures of statistical causality and information-theoretic measures have to say about the problem.
We also make a few intersting observation about biological neural networks.
The task is to predict the existence of edges; and the neural activity is measured through some abra-kadabra spell -- fluorescence imaging.
We are sorry. The title reads "Not so Frequently Asked Questions!" And, this isn't one.
