Managing groups of files in a rule oriented data management system (iRODS)

The iRODS system, created by the San Diego Supercomputing Centre, is a rule oriented data management system that allows the user to create sets of rules to define how the data is to be managed. Each rule corresponds to a particular action or operation (such as checksumming a file) and the system is flexible enough to allow the user to create new rules for new types of operations. The iRODS system can interface to any storage system (provided an iRODS driver is built for that system) and relies on its’ metadata catalogue to provide a virtual file-system that can handle files of any size and type. However, some storage systems (such as tape systems) do not handle small files efficiently and prefer small files to be packaged up (or “bundled”) into larger units. We have developed a system that can bundle small data files of any type into larger units - mounted collections. The system can create collection families and contains its’ own extensible metadata, including metadata on which family the collection belongs to. The mounted collection system can work standalone and is being incorporated into the iRODS system to enhance the systems flexibility to handle small files. In this paper we describe the motivation for creating a mounted collection system, its’ architecture and how it has been incorporated into the iRODS system. We describe different technologies used to create the mounted collection system and provide some performance numbers.

Robust neurofuzzy rule base knowledge extraction and estimation using subspace decomposition combined with regularization and D-optimality

A new robust neurofuzzy model construction algorithm has been introduced for the modeling of a priori unknown dynamical systems from observed finite data sets in the form of a set of fuzzy rules. Based on a Takagi-Sugeno (T-S) inference mechanism a one to one mapping between a fuzzy rule base and a model matrix feature subspace is established. This link enables rule based knowledge to be extracted from matrix subspace to enhance model transparency. In order to achieve maximized model robustness and sparsity, a new robust extended Gram-Schmidt (G-S) method has been introduced via two effective and complementary approaches of regularization and D-optimality experimental design. Model rule bases are decomposed into orthogonal subspaces, so as to enhance model transparency with the capability of interpreting the derived rule base energy level. A locally regularized orthogonal least squares algorithm, combined with a D-optimality used for subspace based rule selection, has been extended for fuzzy rule regularization and subspace based information extraction. By using a weighting for the D-optimality cost function, the entire model construction procedure becomes automatic. Numerical examples are included to demonstrate the effectiveness of the proposed new algorithm.

Rensch's rule in large herbivorous mammals derived from metabolic scaling

Rensch’s rule, which states that the magnitude of sexual size dimorphism tends to increase with increasing body size, has evolved independently in three lineages of large herbivorous mammals: bovids (antelopes), cervids (deer), and macropodids (kangaroos). This pattern can be explained by a model that combines allometry,life-history theory, and energetics. The key features are thatfemale group size increases with increasing body size and that males have evolved under sexual selection to grow large enough to control these groups of females. The model predicts relationships among body size and female group size, male and female age at first breeding,death and growth rates, and energy allocation of males to produce body mass and weapons. Model predictions are well supported by data for these megaherbivores. The model suggests hypotheses for why some other sexually dimorphic taxa, such as primates and pinnipeds(seals and sea lions), do or do not conform to Rensh’s rule.

PMCRI: a parallel modular classification rule induction framework

In a world where massive amounts of data are recorded on a large scale we need data mining technologies to gain knowledge from the data in a reasonable time. The Top Down Induction of Decision Trees (TDIDT) algorithm is a very widely used technology to predict the classification of newly recorded data. However alternative technologies have been derived that often produce better rules but do not scale well on large datasets. Such an alternative to TDIDT is the PrismTCS algorithm. PrismTCS performs particularly well on noisy data but does not scale well on large datasets. In this paper we introduce Prism and investigate its scaling behaviour. We describe how we improved the scalability of the serial version of Prism and investigate its limitations. We then describe our work to overcome these limitations by developing a framework to parallelise algorithms of the Prism family and similar algorithms. We also present the scale up results of a first prototype implementation.