Key points are not available for this paper at this time.
This paper describes a technique for forming approximate, non-deterministic, finite-memory models of behaviour sequences. The models depend on a “memory-length” parameter k which controls their complexity and hence goodness-of-fit lo the given behaviour by constraining them to reproduce correctly all sub-sequences of length k and less in the behaviour. This leads to a simple, incremental, efficient, and surprisingly effective modelling method. Inputs and outputs to the system being modelled are not distinguished but can be intermingled freely in the behaviour sequence, and the extent to which the input process cannot be modelled causally will be reflected in the number of non-deterministic transitions in the model. Examples are included of the use of the technique.
Ian H. Witten (Mon,) studied this question.