The Decoder_Encoder_Model contains codes for building a neural encoder-decoder framework, which estimates the underlying cogntive state using behavioral readout and neural features.
Start with Main_encoder_model_fit.m for fitting the neural encoder model and the neural decoder. The input file contains a matrix of the neural features and behavior decoder model.
The COMPASS StateSpace toolbox contains functions for builiding a model to estimate a hidden state - cognitive state - from behavior such as reaction time and binary decision.