Skip to content

Latest commit

 

History

History
33 lines (19 loc) · 1.82 KB

README.md

File metadata and controls

33 lines (19 loc) · 1.82 KB
Raw

Expected Value of Control Model

  1. Demo simulations

Simulations can be executed with: EVCMood_Simulation1_sampleSubjects (see Matlab file for example parameters)

Plots can be generated with EVCMood_plotEffects(filePath, plotRT) (see Matlab file for example parameters)

  1. Code organization

The EVC model includes the following packages

The class hierarchy of the EVC model itself is set up to allow for multiple implementations of the EVC theory (e.g. a drift-diffusion model (DDM) implementation or a prospective neural network implementation). All specific EVC simulations are implemented in separate classes and inherit parameters from parent classes in order to constrain the parameter space.

  • EVC This package contains classes that implement the general (implementation-independent) computations of the EVC theory (see Shenhav, Botvinick & Cohen, 2013). It also includes classes representing the control signals, the task environment and learning functions.

  • EVC.DDM Classes in this package describe DDM-specific computations such as the retrieval of reaction times and trial outcome probabilities. It also includes classes that provide an interface between implementation-independent model components from the "EVC" package (e.g. the environment and control signals) and the DDM.

  • Simulatons This package includes all simulation files which parameterize the components of a single simulation (e.g. control signals, task environment, etc.). Each simulation file inherits DDM-specific parameters (e.g. default DDM parameters) from the "DDMSim" class which in turn inherits general simulation parameters (e.g. learning rates) from the "Simulation" class. The user may refer to the demo simulation files for more information with respect to how to parameterize a simulation.

Corresponding author:

Sebastian Musslick ([email protected])