This repository contains the data presented in the article "Optimal control of the Wilson-Cowan model
of neural population dynamics".
1. Files labeled as l1_du.pickle,l2_du.pickle,l1_ud.pickle,l2_ud.pickle (or with additional letters/
numbers) relate to the state-switching task. l1 or l2 denote the control task with L1- or L2-cost con-
straints. du or ud denote the down →up or up →down task. The additional _dt denotes the data for the
smaller integration step size dt = 0.01._p1 denotes the data for the parameter set P1. Each file contains
a python dictionary with the following keys:
*"point" – the coordinates (Eext, Iext)or points (1) – (10)
*"cost" – the total cost F=FP+F1or F=FP+F2
*"control" – the optimal control signal
*"state" – the controlled activity
2. Files labeled, e.g., as a_s1_l1.pickle relate to the phase-shifting task. a,b,c,d, or edenote the state-
space point of interest. s1,s2, or s3 denote the simulation duration (T= 140,300,500). l1 or l2 denote
the control task with L1- or L2-cost constraints. The additional _max denotes the data obtained when
initializing the system such that t0occurs at a maximum of the oscillation of the activity variable E.
_pulsetrain denotes the data obtained when initializing the system with a periodic pulse train targeting
the E or I population. _wp denotes the data for a variation of the weight w1._dt denotes the data for the
smaller integration step size dt = 0.001. Each file contains a python dictionary with the following keys:
*"w1" or "w2" – the weights w1or w2.
We computed results for w1,2∈ {0.25,1,4}for a subset of tasks. For most result files, all entries are
None for w1,26= 1.
*"total_cost" – the total cost F=FP+F1or F=FP+F2
*"control" – the optimal control signal
*"state" – the controlled activity
Each dictionary entry is a 41-element list, corresponding to p= 0,0.05π, 0.1π, ..., 1.95π, 2π.
3. The file prc_results.pickle contains the data of the profiles of phase sensitivity at points (A) – (E).
Loading more pages...