Searching for Safety Violations Using Estimation of Distribution Algorithms

Created by W.Langdon from gp-bibliography.bib Revision:1.4192

  author =       "Jan Staunton and John A. Clark",
  title =        "Searching for Safety Violations Using Estimation of
                 Distribution Algorithms",
  booktitle =    "Third International Conference on Software Testing,
                 Verification, and Validation Workshops (ICSTW), 2010",
  year =         "2010",
  pages =        "212--221",
  address =      "Paris",
  month =        apr,
  publisher =    "IEEE",
  note =         "Winner of best student paper",
  keywords =     "genetic algorithms, genetic programming, SBSE, Java
                 PathFinder model checker, concurrent errors, deadlock
                 error, depth-first search, distribution algorithms,
                 model checking, multi-threaded software, safety
                 violation search, Java, concurrency control, formal
                 verification, multi-threading, tree searching",
  DOI =          "doi:10.1109/ICSTW.2010.24",
  abstract =     "Using aspects of model checking to analyse
                 multi-threaded software is a promising method for
                 finding common concurrent errors such as deadlock.
                 Traditional model checking tools exhaustively search
                 the state space of a concurrent system in order to find
                 faults. Unfortunately, model checking suffers from the
                 state space explosion problem, limiting the
                 applicability of the approach to commercial software.
                 Metaheuristic search mechanisms have been used in an
                 attempt to overcome this issue with good results.
                 Techniques such as Genetic Algorithms (GAs) and
                 Estimation of Distribution Algorithms (EDAs) focus the
                 search of the state space on areas that are more likely
                 to contain errors. In this work, a novel EDA-based
                 approach to exploring the state space of a model is
                 outlined. Experiments are performed on an
                 implementation using the Java PathFinder (JPF) model
                 checker and the ECJ toolkit. The EDA-based approach is
                 shown to perform well against standard search
                 procedures such as depth-first search, whilst also
                 outperforming random search on a benchmark problem. On
                 larger problems, the EDA is shown to be the only
                 effective technique of those compared.",
  notes =        "SBST 2010 also known as \cite{5463650}",

Genetic Programming entries for Jan Staunton John A Clark