doi:10.3850/978-981-08-5118-7_plenary2
Data Mining Applications in the Automotive Industry
Rudolf Kruse, Matthias Steinbrecher and Christian Moewes
Computational Intelligence Group, Department of Knowledge and Language Processing
Faculty of Computer Science, Otto-von-Guericke University of Magdeburg
Universitätsplatz 2, D-39106 Magdeburg, Germany
ABSTRACT
Designing and assembling automobiles is a complex task which has to be accomplished
in ever shorter cycles. However, customers have increasing desires w. r. t. reliability, durability and
comfort. In order to cope with these conflicting constraints it is indispensable to employ tools that
greatly simplify the analysis of data that is collected during all car lifecycle stages. We will present
methods for pattern discovery tasks for the development stage, the manufacturing and planning
stage as well as for maintenance and aftercare. The first approach will reinterpret a Bayesian
network to induce association rules which are then visualized to find interesting patterns. The
second part will use Markov networks to model the interdependencies related to the planning task
when assembling a vehicle. The last part deals with finding recurring patterns in time series used
for adjusting simulation parameters.
Back to TOC
FULL TEXT(PDF)
|