EXAMPLES OF FLEXIBLE BAYESIAN REGRESSION AND CLASSIFICATION MODELS BASED ON NEURAL NETWORKS AND GAUSSIAN PROCESSES Neural networks and Gaussian processes can be used to to predict a target value (or values) from a set of input values. When the target is real-valued, this is known as "regression". When the target takes on values from a finite set, it is known as "classification". Simple examples of both sorts are discussed here. Some more complex classification models for images are discussed in Ex-image.doc. The data and command files for these examples are in the "ex-netgp" directory.