Classification tree chaining in data analysis
Understanding of discrete event simulation systems is a difficult task.
Our approach involves consultation
with a domain expert, and the use of discrete event simulation model and
machine learning as tools for the intelligent analysis of simulated system.
Current methods for the analysis and interpretation of such systems are
restricted to statistical techniques that say much about the reliability of
an output, but little about the output inter connectivity.
The objective of our
work is to improve the ability to interpret the model to the level of
explanation that might loosely be described as ``How the simulated system
works''. The new interpretation techniques are based on
machine learning that includes classification tree chaining.
``How the simulated system works'' is important for a decision maker's
understanding of
the system in terms of relationships between the various parameters, the
utilisation of resources and the location of (potential) bottlenecks. This
sort of understanding is important when managing queuing problems in
production systems, or when planning a new system. The output from a
discrete event simulator was processed as machine learning problem.
In these experiments, the simulation output was interpreted by
attribute-based learning system Magnus Assistant.
The knowledge gained from the machine learning tools was evaluated
by the domain expert, who considered that this knowledge contributed
to an understanding of the system, and how it might profitably be modified.