Three courses are currently under development. The first course is a general introduction to simulating biologically
based neural networks and utilizes MatLab rather than NeuroJet. The second course teaches, through examples, the NeuroJet scripting language. A novice can run the tutorial scripts. The third course is under development and teaches users how to simulate on multinode parallel computers with MPI technology.
Introduction to Simulating Biologically Based Neural Networks
Three labs from the introductory course are available below in Microsoft Word format. The course assumes that the user has access to MatLab software. The available materials are usually presented in an undergraduate course at the University of Virginia, but the material should be accessable to the ambitious high school student and can serve as a good starting point for the interested graduate student. The first lab familiarizes the student with the MatLab environment, discusses the importance of random numbers, introduces key elementary statistical ideas and explains normalization of vectors. The second lab explains how artificial neurons can act as decision-making and pattern recognition devices. Lab 2 also instructs the student on the interactions of threshold and input excitation; it also demonstrates the limitations of linear discrimination as performed by one neuron. The third lab introduces moving averagers and their behavior.
Laboratory 1 : Introductory Concepts : MatLab, Random Numbers and Neurons
Laboratory 2 : Neuronal Pattern Recognition
Laboratory 3 : Moving Averagers
The tutorials for NeuroJet introduce the user to the sequence learning paradigm and show the new user examples of the powerful NeuroJet scripting language. It also demonstrates the use of MatLab for analysis and visualization of data from NeuroJet.