The marine currents around the Pacific Northwest are researched by lots of scholars. The methods of the on-site observation, such as the shipping measurements and data buoys, were applied by researcher in the past. However, it is difficult to get comprehensive information on space by early methods, and it is often limited by the manpower and funding. With computer technology advances, numerical models get improving the accuracy and resolution. Therefore, development of the numerical models also becomes one of the methods to understand the ocean. There are many current models and large databases which are developed. However, few studies combine the two. Therefore, this study adopts the simulated results of POLCOMS (Proudman Oceanographic Laboratory Coastal-Ocean Modeling System) in 2002 near the northwest Pacific (9.9¢X N ~ 41.9¢X N, 104.9¢X E ~ 139.9¢X E) to analyze some situations. The result can be compared with the measured data, and test the model. In order to evaluate the accuracy of the results of the model, the simulated tracks can be calculated by the result of the model and the initial location of the measured data. However, the resolutions of time and space are only one day and 7.5 degree. Hence, the methods of linear interpolation and bilinear interpolation are applied to interpolate the model result in the time dimension and space dimension. By using the latitude and longitude of the initial points, the azimuth, and the geographic distance, the simulated tracks can be calculated. Therefore, the results of POLCOMS can be confirmed by global drifter data from NOAA/AOML. According to the results of evaluating, the simulation trends of the seasons are similar to measured data. It can prove that this model has a certain degree of accuracy. However, simulation and the measured data can not be completely similar when abnormal weather phenomenon occurs. It is because the model was driven by the average wind field. There are suggestions to make the model complete by adding the weather information during the periods of the typhoons. Besides, the model can simulate the temperature and salinity of the ocean. Increasing the comparison of them will make the model integrated in the future, and obtain more correct information of the flow fields around the Pacific Northwest.

The objective of this thesis is to design and implement a speech recognition system for Mandarin names and phrases. This system utilizes Mel frequency cepstral coefficients, hidden Markov model and lexicon search strategy to select the phrase candidates. The experimental results indicate that for the speaker dependent case, a strategy incorporating overlapping frames and hybrid training can result in an improvement of 4%, 5%, 4% and 2% on the recognition rate for the Mandarin name, two-word, three-word and four-word phrase recognition systems respectively. Under Redhat Linux 9.0 operating system, any Mandarin name or phrase can be recognized within 2 seconds by a computer with Intel Celeron 2.4 GHz CPU.

Euclidean distance transformation is a fundamental technique in image understanding and computer vision. Some important characteristics in image analysis such as skeleton and object boundary are based upon the distance transformation computation. In this thesis, we compare our method of computing Euclidean distance transformations with the method of Chamfer distance transformation. Our method is faster and more accurate than the Chamfer method. The boundary detection is an interesting and challenging task in computer vision. We integrate distance transform, watershed transform and active contour model to achieve boundary detection. Our method can successfully separate the touching objects, so as to facilitate the subsequent image processing for obtaining the geometric, and texture characteristics of objects. These features are useful for further medical images applications.

Inaug. Diss.--Bonn. / Bibliography: p. [71].

Thesis (Ph. D.)--University of Texas at Austin, 2003. / Supervisor: Linda E. Reichl. Vita. Includes bibliographical references. Also available from UMI.

This research was conducted on model-based reasoning and its prevalence in CSCOPE curriculum. Communications with seven CSCOPE representatives out of twenty regions revealed that CSCOPE is simply a name, not an acronym. The primary focus of CSCOPE is to impact instructional practices in the classroom to improve student performance. This report discusses the history of CSCOPE, its framework, and its exemplar lessons. It also looks at model-based reasoning, taxonomy of models, and model-eliciting activities. The research also aims to determine if the exemplar lessons in CSCOPE can be classified as model-based. / text

A passenger mix model (PMM) is used by airlines to find out how many passengers will fly on a fleet schedule. There are numerous ways of modeling passenger mix models and this report studies a share based passenger mix model, proposed by Sabre, and tests its efficacy against a deterministic linear program (DLP) passenger mix model. A DLP passenger mix model cannot recapture spilled passengers and requires iterations of the same model to recapture passengers. In order to eliminate the iterative nature of the DLP model Sabre proposed a new model which eliminates iterations for recapturing passengers. This report studies the proposed share based passenger mix model and compares it with the DLP model in terms of traffic allocation and speed of solution. It is found that the share based model allocates traffic in the same manner as the tried and tested DLP model. / text

This thesis centers around the pricing and risk-return tradeoff of credit and equity derivatives. The first essay studies the pricing in the CDS Index (CDX) tranche market, and whether these instruments have been reasonably priced and integrated within the financial market generally, both before and during the financial crisis. We first design a procedure to value CDO tranches using an intensity-based model which falls into the affine model class. The CDX tranche spreads are efficiently explained by a three-factor version of this model, before and during the crisis period. We then construct tradable CDX tranche portfolios, representing the three default intensity factors. These portfolios capture the same exposure as the S&P 500 index optionmarket, to a market crash. We regress these CDX factors against the underlying index, the volatility factor, and the smirk factor, extracted from the index option returns, and against the Fama-French market, size and book-to-market factors. We finally argue that the CDX spreads are integrated in the financial market, and their issuers have not made excess returns. The second essay explores the specifications of jumps for modeling stock price dynamics and cross-sectional option prices. We exploit a long sample of about 16 years of S&P500 returns and option prices for model estimation. We explicitly impose the time-series consistency when jointly fitting the return and option series. We specify a separate jump intensity process which affords a distinct source of uncertainty and persistence level from the volatility process. Our overall conclusion is that simultaneous jumps in return and volatility are helpful in fitting the return, volatility and jump intensity time series, while time-varying jump intensities improve the cross-section fit of the option prices. In the formulation with time-varying jump intensity, both the mean jump size and standard deviation of jump size premia are strengthened. Our MCMC approach to estimate the models is appropriate, because it has been found to be powerful by other authors, and it is suitable for dealing with jumps. To the best of our knowledge, our study provides the the most comprehensive application of the MCMC technique to option pricing in affine jump-diffusion models. / published_or_final_version / Economics and Finance / Doctoral / Doctor of Philosophy

Not available / text

This thesis presents the results of an investigation into the application of statistical energy analysis (SEA) to predict structure-borne noise transmission in ship structures. The first three chapters introduce the problems of noise and vibration in ships; the previous research on the application of SEA to ships; the basic theory of SEA and the experimental measurement techniques and procedures used to gather data The main body of this thesis presents a wave transmission model for the hull frame joint which is commonly encountered on the hull, bulkheads and deck plates of ship structures. The wave model allows the transmission coefficients to be calculated for hull frame joints which can be used in the coupling loss factor equations of SEA models. The joint model has been verified against measured data taken on a simple two subsystem single joint laboratory structures and a large complex 38 plate test structure with multiple joints intended to represent a 1/10' scale model of a hull section. In addition to the laboratory structures, the SEA modelling of sections of a ship is presented for a large ribbed deck plate, a section of the ship superstructure and a section of the ships hull. The results from the SEA models are compared with measured attenuation data taken on the respective ship sections. A large amount of damping data has been gathered on the test and ship structures and an equation for the internal steel based on data gathered by other researchers has been verified. It has been shown in this thesis that SEA can be applied to ships. Better agreement is found with real structures in contrast to the poor results presented for SEA when applied to simple one dimensional structures. The level of detail of the model is important as a coarse model yields better predictions of vibration level. As with all models the results are sensitive to the damping level and it is necessary to include bending, longitudinal and transverse wave types in any SEA model to obtain the best prediction. It was also found that the flange plates can be neglected from the frame joint model without compromising the accuracy.