Using cellular automata to predict reliability of modules

Software development is complex process that incorporates many different aspects and is therefore hard to control. For that reason many different metrics are used to capture different aspects of software artefact. On other hand, data-mining from such database should also contain different aspects and try to express nature of software development process. For that reason we use an ensemble of classifier that have the ability to boost classification accuracy comparing to single classifiers and are a commonly used method in the field of machine learning. However in some cases ensemble construction algorithms do not improve the classification accuracy. Mostly ensembles are constructed using specific machine learning method or a combination of methods, the drawback being that the combination of methods or selection of the appropriate method for a specific problem must be made by the user. To overcome this problem we invented a novel approach where ensemble of classifiers is constructed by a self-organizing system applying cellular automata (CA).