No description available.

Westphalian non-marine bivalves in Nova Scotia inhabited fresh to brackish temporary lakes. A series of lacustrine facies are identified on the basis of the faunas. Myalinid genera including the probably pseudoplanktonic Curvirimula, and Naiadites dominate the middle Westphalian A to early Westphalian B. Carbonicola occurs rarely in the late Namurian to early Westphalian A and is the only reported Anthracosiid. Anthraconauta dominates the upper Westphalian C to early Stephanian occurring in a wide range of sediments. The elongate Anthraconaia arenacea group range through the Westphalian D and occur in coarser lithologies. Small-shell Anthraconaia inhabited ephemeral lake margins. Cluster and principal components analysis are used to classify the faunas. The method requires the measurement of a small number of linear and relatively uncorrelated morphological variables and allows the construction of reproducable pictographs. Three new morphospecies are described. Principal components analysis is used to study morphological variation with respect to palaeoenvironment. Consistent morphological trends occur in Naiadites and Anthraconauta and are considered to have been the product of spat selection and phenotypic changes. With increased turbulence and sedimentation rates, the shell's centre of gravity is shifted towards the anter-umbonal region and the bivalves adopt a semi-infaunal mode of life thus, the morphological changes offer increased stability. Evolution was inhibited by the absence of marine incursions and progressive environmental change. Such palaeoenvironmental changes that did occur were local. Competition may have been of evolutionary importance in younger communities of Anthraconaia. The bivalves are of limited stratigraphical use as compared to the British faunas. In conjuction with micro- and macrofloral work, the Namurian/Westphalian A and Westphalian C/D boundaries are lowered and a possible Cantabrian stage recognised. A correlation with the British Chronozones is attempted.

No description available.

Binary survey data from the Austrian National Guest Survey conducted in the summer season of 1997 were used to identify behavioral market segments on the basis of vacation activity information. Bagged clustering overcomes a number of difficulties typically encountered when partitioning large binary data sets: The partitions have greater structural stability over repetitions of the algorithm and the question of the "correct" number of clusters is less important because of the hierarchical step of the cluster analysis. Finally, the bootstrap part of the algorithm provides means for assessing and visualizing segment stability for each input variable. (author's abstract) / Series: Report Series SFB "Adaptive Information Systems and Modelling in Economics and Management Science"

No description available.

Strategic information systems planning (SISP) remains a primary concern for many organisations. It is reportedly a top concern of not only the Information Systems (IS) executives but also of many chief executives. Recently, calls have been madefor better understanding of strategic planning in general, including SISP, and especially for studies of actual planning behaviour in organisations. As doubts continue to be raised about the payoff ofIT, it is important to examine the effectiveness ofmany accepted IS management practices such as SISP. Accordingly, researchers have investigated SISP practice and proposed both formal methods and principles of good practice. SISP cannot, however, be understood by considering formal methods alone. Yhe processes of planning and the implementation of plans are equally important However, there have been very few investigations of these phenomena. More surprisingly, few studies have been undertaken to identify and understand: (]) the approaches to SISP adopted in practice, (2) the relationship between SISP approaches and its ultimate success, (3) the influence of organisational and technological context on planning success. Yhis study helps to fill this vacuum in the important research area. A formal survey of IS Directors, which generated two hundred and ninety two, responses was initiated to empirically investigate these issues. The categorisation of SISP approaches adopted by companies was accomplished by performing cluster analysis, based on chosen IS planning dimensions, namely., comprehensiveness, participation, formalisation, flow, focus, frequency, alignment, ownership, IS benefit and plan implementation. Using this multivariate technique, four unique and stable groups were identified, namely: (1) organisational, (2) business-led, (3) administrative and (4) formal, This derived taxonomy was thoroughly validated to ensure that it is exhaustive, mutually exclusive, stable and consistent. Within this study,S ISPs uccessh as been conceptualisedb asedo n the extento f thefollowing measures: (1) alignment, (2) analysis, (3) co-operation, (4) implementation, (5) capabilities,( 6) satisfactiona nd (7) contribution. Thes tudyh as also conceptualisedS ISP success in terms of an aggregate score. It is found that each approach, with its different characteristics, has a different likelihood of success. Thefindings of this research suggest that the Organisational approach, which is characterised by high levels of alignment, comprehensivenespsa, rticipation,f requency,i nfusion benefita ndp lan implementationi,s by far the most successful. Conversely, the Administrative approach which has low levels of comprehensivenessfr, equency, participation and alignment is the least successful. Interestingly,t his study also suggeststh at the Organisationala pproach deliversa high level ofSISP successir respectiveo fthe organisationala nd technologicalc ontexto f the company. In summary, this study has extended our understanding of SISP processes and has provided useful insightsfor IS executives and top management in general, in implementing SISP within their organisations.

We propose a novel framework based on Bayesian principles for validating clusterings and present efficient algorithms for use with centroid or exemplar based clustering solutions. Our framework treats the data as fixed and introduces perturbations into the clustering procedure. In our algorithms, we scale the distances between points by a random variable whose distribution is tuned against a baseline null dataset. The random variable is integrated out, yielding a soft assignment matrix that gives the behavior under perturbation of the points relative to each of the clusters. From this soft assignment matrix, we are able to visualize inter-cluster behavior, rank clusters, and give a scalar index of the the clustering stability. In a large test on synthetic data, our method matches or outperforms other leading methods at predicting the correct number of clusters. We also present a theoretical analysis of our approach, which suggests that it is useful for high dimensional data.

Shared memory systems, such as SMP and ccNUMA topologies, simplify programming and administration. On the other hand, systems without hardware support for shared memory, such as clusters of commodity workstations, are commonly used due to cost and flexibility considerations. In this thesis, virtualisation is proposed as a technique that can bridge the gap between these architectures. The resulting system, vNUMA, is a hypervisor with a unique feature: it provides the illusion of shared memory across separate nodes on a fast network. This allows a cluster of workstations to be transformed into a single shared memory multiprocessor, supporting existing operating systems and applications. Such an approach could also have applications for emerging highly-parallel architectures, allowing a shared memory programming model to be retained while reducing hardware complexity. To build such a system, it is necessary to meld both a high-performance hypervisor and a high-performance distributed shared memory (DSM) system. This thesis addresses the challenges inherent in both of these tasks. First, designing an efficient hypervisor layer is considered; since vNUMA is implemented on the Itanium processor architecture, this is with particular reference to Itanium processor virtualisation. Then, novel DSM protocols are developed that allow SMP consistency models to be reproduced while providing better performance than a simple atomically-consistent DSM system. Finally, the system is evaluated, proving that it can provide good performance and compelling advantages for a variety of applications.

This study uses clusters formed in a UHV-compatible cluster apparatus, which was built and commissioned during this thesis. The design and operation of the cluster deposition system is described. This system is optimised for high clus- ter fluxes and for the production of cluster assembled nanoscale devices. One key feature of the system is a high degree of flexibility, including interchangeable sputtering and inert gas aggregation sources, and two kinds of mass spectrome- ter, which allow both characterisation of the cluster size distribution and deposi- tion of mass-selected clusters. Another key feature is that clusters are deposited onto electrically contacted lithographically defined devices mounted on an UHV- compatible cryostat cold finger, allowing deposition at room temperature as well as at cryogenic and at elevated temperatures. The electrically contacted nanoscale cluster devices were fabricated using a novel template technique. Hereby, clusters are placed between two electrodes separated only by ∼100 nm. The width of the cluster ensemble is in the order of a few cluster diameters, which means that the assembled clusters form a cluster wire bridging the electrode separation. During this thesis, the design and layout has been optimised to be able to measure electrical properties of the cluster devices and in particular to investigate the interaction between the cluster ensemble and the contact electrodes. In-situ electrical characterisation of cluster assembled nanoscale devices are performed in the temperature range 4.2 K to 375 K. The samples are provided with a backgate, which in principle allows modification of the conduction through the cluster ensemble by applying a gate voltage. However, no change in conduc- tion with changes in gate voltages was seen. The main focus of the electrical measurements is on the current voltage char- acteristics. It was noticed that the nanoscale bismuth (and antimony) cluster devices exhibited non-linear current voltage characteristics, which were in stark contrast to the linear current voltage characteristics measured for cluster films previously. Investigations into the causes of this non-linearity suggests that tun- nelling conduction occurs between the cluster ensemble (wire) and the contact electrodes. The non-linear current voltage characteristics were fitted using three models of tunnelling conduction and appear to be best fitted using a model in- volving fluctuation-assisted tunnelling through barriers of different heights. Further, measurements of the temperature dependent resistance are performed showing an increase of resistance with decreasing temperature for bismuth and antimony assembled cluster devices. The temperature dependence of bismuth as- sembled cluster wires can be explained by the decrease of the carrier concentration in bismuth for decreasing temperature. Annealing of the cluster ensemble and the cluster contact connection resulted in an increase in conduction. This increase of conduction can be explained due to the current flow through the cluster wire. Locally, at the bottlenecks, the current flow causes resistive heating and subsequently coalescence of two (or more) clusters.

Thesis (Ph.D.)--Tufts University, 2001. / Adviser: Alva Couch. Submitted to the Dept. of Computer Science. Includes bibliographical references (leaves 144-148). Access restricted to members of the Tufts University community. Also available via the World Wide Web;