iDEA: Drexel E-repository and Archives >
Drexel Academic Community >
College of Information Science and Technology >
Research Day Posters (IST) >
Automated configuration management for computational intelligence systems
Please use this identifier to cite or link to this item:
|Title: ||Automated configuration management for computational intelligence systems|
|Authors: ||Proctor, Jason M.|
Weber, Rosina O.
|Keywords: ||Artificial intelligence;Expert systems;Problem solving|
|Issue Date: ||6-Sep-2005|
|Series/Report no.: ||IST Research Day 2005 posters;no. 68|
|Abstract: ||The configuration of a computational intelligence (CI) method is responsible for its tolerance, flexibility, and accuracy in solving complex problems. Configuring a system that implements these robust and often non-deterministic methods is also a complex problem, and researchers usually start with suggestions from the literature then apply a mix of intuition, trial-and-error, and luck. Further complicating the issue is that the same CI system may be able to solve radically different problems which require different configurations. Our work is aimed at improving the effectiveness of these systems by systematically finding a suitable set of values for configuration parameters to solve each problem, and managing the knowledge gained along the way to make it easier to solve future problems using the same system. We validate our methods in a system that uses an artificial neural network to aid software testing, and show potential future applications in biological systems.|
|Appears in Collections:||Research Day Posters (IST)|
Items in iDEA are protected by copyright, with all rights reserved, unless otherwise indicated.