Nonlinear Auto-associative Neural Network

Hopfield model is one example of the neurodynamics, which is also nonlinear, and is auto-associative (in the sense that input is also output).  In Hofield model, the basic degree of freedom is discrete.  If the variables can take continuous real values, then the governing equations should be coupled differential equations.  The theory of nonlinear dynamics can be used.  Such concepts as fixed points, linear stability analysis and chaos are important.  This is outside the scope of this course.  In the next section, we discuss a particular neurodynamic model, the "brain-state-in-box" model, which finishes up this part of the courses.