published_or_final_version / Applied Geosciences / Master / Master of Science

The aim of this project is to study the impact of future sea level rise on coastal development in Hong Kong and to review current marine design work standard of Port Work Design Manual. Projection of sea level rise to 2060 and 2099 from past tide gauge records has been carried out in this study. The recent redevelopment area, Kai Tak (Old airport), is chosen as study area. Data has been obtained from Hong Kong Observatory, IPCC Fourth Assessment Report (2007), Lands Department and Civil Engineering and Development Department to reconstruct and project the past and future sea level change in Hong Kong respectively. Global and local sea level are expected to rise in the 20th century from past studies. The increasing acceleration of sea level rise in my projection has an significant implication to Kai Tak area that existing formation level of reclaimed land and current marine design work standard for redevelopment may not be adequate to resist coastal flooding and inundation associated with sea level change in the near future. By 2099, Kai Tak area may be flooded in extreme cases, which is an issue that upgrading existing marine structure and reviewing current marine design standard needs immediate attention. / published_or_final_version / Applied Geosciences / Master / Master of Science

This dissertation investigates the High Level Architecture (HLA) as a possible distributed simulation framework for transportation systems. The HLA is an object-oriented approach to distributed simulations developed by the Department of Defense (DoD) to handle the issues of reuse and interoperability of simulations. The research objectives are as follows: (1) determine the feasibility of making existing traffic management simulation environments HLA compliant; (2) evaluate the usability of existing HLA support software in the transportation arena; (3) determine the usability of methods developed by the military to test for HLA compliance on traffic simulation models; and (4) examine the possibility of using the HLA to create Internet-based virtual environments for transportation research. These objectives were achieved in part via the development of a distributed simulation environment using the HLA. Two independent traffic simulation models (federates) comprised the environment (federation). A CORSIM federate models a freeway feeder road with an on-ramp while an Arena federate models a tollbooth exchange.

Disk performance for random access fares significantly worse compared to sequential access. Time required to transfer random blocks to or from disk is dominated by seeking and rotational delay. To improve the throughput and reduce the latency, one can apply techniques to increase the sequentiality of disk accesses, such as block rearrangement and replication. We introduce an approach to improve read performance by replicating blocks into file system free space at the block level. This makes the replication module independent of the file system and therefore easier to implement and verify. A solution that requires no changes to the file system is also easier to adopt. Supporting a new file system is a matter of writing a user-space component that understands its free block data structures. We implemented a prototype as a stacked device driver for Linux and evaluated its performance on a number of workloads.

No description available.

No description available.

No description available.

Les transformations dans la structure phonétique ont toujours accusé une tendance prononcée à l'articulation antérieure et labialisée ce qui aboutit à l'enrichissement considérable des consonnes prélinguales ( affriquées > consonnes chuintantes et sifflantes, semi-consonnes ) et des voyelles antérieures; celles-ci comportant deux oppositions importantes: ouvertes/fermées, labialisées/non labialisées. La quatrième série d'oppositions est présentée par l'opposition voyelle orale/voyelle nasale, constituée vers le XVIe s. La différenciation qualitative est à toutes les époques la caractéristique essentielle du vocalisme français: la durée vocalique peut aller de pair ou bien porter un caractère phonétique complémentaire. Les oppositions phonématiques parmi les consonnes sont moins générales que parmi les voyelles. Elles réunis¬sent les groupes de consonnes plus ou moins nombreux, laissant en marge plusieurs autres. Ainsi, l'opposition « sour¬de-sonore » ( srd-snr ) se manifeste seulement parmi les consonnes-bruits, les sonantes étant sonores par excellence. Une autre opposition « orale-nasale » frappe seulement les occlusives, les constrictives étant toutes orales.

The overall aim of this thesis is to uncover the key determinants of innovation in New Zealand firms and consider some of their likely effects. In order to provide a broad perspective on New Zealand’s local innovation processes, a mixed method approach combining both quantitative and qualitative analysis was adopted to allow analysis of both empirical data and case study data. The quantitative part of analysis utilises the unique dataset developed by Statistics New Zealand, namely the prototype Longitudinal Business Database (LBD), and the qualitative analysis includes four in-depth company case studies which complement the regression analyses by uncovering the key patterns of innovation behaviour at the firm level. In summary, a number of conclusions have been drawn from the research. Firstly, firms experience considerably smaller positive size effect because of New Zealand’s unique firm demographics, and the small size has limited individual firm’s innovation opportunities. Secondly, firms’ ability to develop new technologies directly influences their innovative ability, which is highly dependent on the availability of funds and skills. Lastly, innovation in New Zealand has a very strong market focus, while technology suppliers such as universities and Crown Research Institutes only have a limited role in selected industries.

One of the most active research areas in computer vision is image classification. Although there have been many research efforts in this area, it remains a difficult problem, especially when the number of categories is large. Most of the previous work in image classification uses low-level image features. We believe low-level features ignore a lot of the semantic structures of the image classes. In this thesis, we go beyond simple low-level features and propose new approaches for constructing mid-level visual features for image classification. We represent an image using the outputs of a collection of binary classifiers. These binary classifiers are trained to differentiate pairs of object classes in an object hierarchy. Our feature representations implicitly capture the hierarchical structure in object classes. We show that our proposed approach outperforms other baseline methods in image classification.