GBLM-SMOOTH is a model, implemented in MATLAB, that can track long-term and short-term synaptic plasticity simultaneously. This repository contains four folders.
Three sub-folders are included:
(i) "diagnose": tools to evaluate fitting.
(ii) "modelFit": implementation of the model. The basic model (without tuning Q in adaptive smoothing) starts from smooth_gblm.m. The Q-tuned versions are generally called tune_smooth_gblm_xd_xxxx.m. "xd"" means whether it is 2D optimization or 1D approximation. The last four letters denote optimization methods, i.e. "_grad" for gradient descent and "_grid" for grid search.
(iii) "simulation": code for generating neural spikes.
This folder provides code for generating all figures in the paper. Generally, "_caseDemo" denotes code for simulations and model fitting, while "_caseAnalysis" denotes code for plotting.
It contains helper functions for simulations and model fitting.
This folder contains all the plots used in the paper.