Features

Last Updated February 22, 2005

No part of this web site may be reproduced except by written permission of the copyright holder

Simulation

• Dynamical system models for maximum flexibility

• Biologically-realistic, systems-level networks

• Simulates Hebbian learning, reinforcement learning, supervised learning

• Simulate effects of multiple neurotransmitter and receptor types

• Includes pre-defined activation and weight governing equations

• Governing equations for weight, activation are independent – can mix and match

• Built-in scripting language allows user-defined custom equations

• Rate coding or spiking cells (integrate and fire)

• Weights and activations update continuously

Data Fitting

• Fit multiple data types in one simulation – behavior, fMRI, neurophysiology, and more

• Powerful gradient descent algorithm automatically fits model to data, regardless of governing equation choice or complexity of data set

• Dissociation of learning and parameter fitting allows learning-related data to be fit

Ease of Use

• PDP++ graphical user interface

• Parallel processing support for data fitting