Assume a unidirectional circuit that transforms information. If one records the activity from the middle and end of the circuit they will end up with two times series. My goal is to predict the formula that governs this transformation or at the very least a model that can predict B given A. I've looked at VAR, Granger causality, and Recurrent neural Networks, but need a bit of guidance to be pointed in the right direction. I thought of using the B data I have as training data, but I'm also interested in accuracy testing methods. I can include data if needed.
take a look at the work being done in the hippocampus of decoding location from place cell activity. instead of decoding location you'd be decoding firing in area B.
Here are some references:
Brown EN, Frank LM, Tang D, Quirk MC, Wilson MA. A statistical paradigm for neural spike train decoding applied to position prediction from ensemble firing patterns of rat hippocampal place cells. J Neurosci. 1998; 18:7411–7425. [PubMed: 9736661]
F. Kloosterman. Analysis of Hippocampal Memory Replay Using Neural Population Decoding, in Neuromethods (2012), pp. 259–282.