{\rtf1\ansi\ansicpg1252\cocoartf1504\cocoasubrtf830
{\fonttbl\f0\fswiss\fcharset0 Helvetica;}
{\colortbl;\red255\green255\blue255;\red25\green25\blue25;}
{\*\expandedcolortbl;;\cssrgb\c12941\c12941\c12941;}
\paperw11900\paperh16840\margl1440\margr1440\vieww10800\viewh8400\viewkind0
\pard\tx566\tx1133\tx1700\tx2267\tx2834\tx3401\tx3968\tx4535\tx5102\tx5669\tx6236\tx6803\pardirnatural\partightenfactor0

\f0\fs24 \cf0 The present folder contains all the required MATLAB scripts and functions to run the sparse logistic regression (SLR) framework introduced in Bolton et al.\'92s \'93\cf2 \expnd0\expndtw0\kerning0
Sparse coupled logistic regression to estimate co-activation and modulatory influences of brain regions\'94 (presently submitted to Journal of Neuroengineering).\
\
The \'93RSCHMM_Simulations_Figure*.m\'94 scripts contain the codes to generate the simulations presented in Figures 2 to 5 from the paper.\
\
The \'93RSCHMM_Computations*.m\'94 scripts shows all the codes to perform the computations of the SLR framework, step by step, and also displays some example plotting utilities. Looking at this script should be enough to get a good sense of all the steps of the SLR pipeline.\
\
The \'93LogisticRegressionFramework\'94 folder contains all the MATLAB functions required to summon the SLR framework and associated utilities. Within this folder, the \'93Framework\'94 sub-folder contains the SLR routines themselves. The \'93Miscellaneous\'94 and \'93OlderImplementations\'94 folders in there are not required to run the latest version of the pipeline.\
The \'93Plotting\'94 sub-folder contains utilities to graphically investigate some of the model\'92s outputs.\
The \'93PostProcessing\'94 sub-folder contains the codes that perform further computations from the coefficients themselves (e.g., converting them into probabilistic couplings between pairs of regions).\
The \'93Simulations\'94 sub-folder contains the codes needed to run the simulations presented in the article. The \'93Miscellaneous\'94 folder in there can be discarded, as it contains unused alternative attempts at simulating.\
\
The \'93Script_OtherApproaches.m\'94 script shows how to extract co-activation or causal coefficient matrices with the four approaches compared to our SLR framework: the graphical lasso, a point-process analysis, a multivariate autoregressive model, or a cross-spectral density-based approach.\
\
The \'93OtherApproaches\'94 folder contains the MATLAB functions used to perform computations with other approaches compared to the SLR framework in the paper\'92s Figure 4.}