Thesis (Ph. D.)--Ohio State University, 1983. / Includes bibliographical references (leaves 136-146). Available online via OhioLINK's ETD Center

The crust is the outer-most layer of the earth with thickness up to 80 km. Massive seismic waveform data have enabled imaging fine crustal structures with the aid of new imaging techniques. In this thesis, I develop seismic imaging techniques to take full advantage of the expanding dataset as well as apply the imaging techniques to understand crustal seismic structures. First, I apply receiver function techniques to image the crustal thickness and average Vp/Vs in Northeast China. I found an uplifted Moho in eastern flank of the Songliao Basin and the Changbaishan region and suggest that dynamic mantle upwelling might be the cause of the observed uplift. With accumulated waveform data available, it becomes possible to extract more subtle structural information from receiver function. Second, I develop a new technique to robustly estimate seismic azimuthal anisotropy with radial and transverse receiver functions. I apply this technique to estimate the crustal anisotropy in Southeast Yunnan region and found that the significant crustal anisotropy may be caused by lower crust flow in this region. Full-wave based imaging techniques such as reverse time migration and full-wave inversion does not assume flat interfaces or infinite frequency rays as that the receiver function techniques do and are desirable in imaging more complex crustal structures. However, their high computational cost is one of the issues that prevent their practical applications. In the last part, I developed an effective waveform modeling technique to efficiently simulate wave propagation in acoustic media. With this novel modeling technique, the full-wave based imaging techniques are accelerated by a factor up to 400%.

The role played by the protein UnkG from the plant growth-promoting bacterium Pseudomonas putida UW4 in the ability of the bacterium to facilitate plant growth was studied. Previous work showed that over-expressing UnkG decreased the ability of P. putida UW4 to facilitate plant growth. In contrast, an unkG knock-out mutant of P. putida UW4 displayed an increased ability to promote plant growth. Various biological activities of P. putida UW4, P. putida UW4/pETP and P. putida UW4/pETP-unkG have been compared. Thus, the growth curves were measured; the Biolog™ system was used to test the ability of these strains to utilize various carbon sources; the strains were observed by scanning electron microscopy to assess their relative cell sizes; biochemical assays were conducted to quantify 3-indoleacetic acid production and to measure the enzymatic activity of 1-aminocyclopropane-1-carboxylate (ACC) deaminase; proteome-level changes of P. putida UW4/pETP and P. putida UW4/pETP-unkG were profiled using two-dimensional difference in-gel electrophoresis (DIGE), followed by mass spectrometry identification of the altered proteins. After running DIGE, sixteen altered proteins were identified and their possible roles in the interactions between the bacterium and plants were discussed. Based on the preliminary results, we hypothesize that 1) UnkG may be detrimental to plant growth; 2) UnkG may negatively regulate a number of key cellular functions in a general way related to the energy balance of the bacterium.

We define the phylogenetic model of a trivalent graph as a generalization of a binary symmetric model of a trivalent phylogenetic tree. If the underlining graph is a tree, the model has a parametrization that can be expressed in terms of the tree. The model is always a polarized projective toric variety. Equivalently, it is a projective spectrum of a semigroup ring. We describe explicitly the generators of this projective coordinate ring for graphs with at most one cycle. We prove that models of graphs with the same topological invariants are deformation equivalent and share the same Hilbert function. We also provide an algorithm to compute the Hilbert function, which uses the structure of the graph as a sum of elementary ones. Also, this Hilbert function of phylogenetic model of a graph with g cycles is meaningful for the theory of connections on a Riemann surface of genus g.

The dielectric waveguide component has become a mature industry in these days in making it and understanding how it works. There are many theoretical and numerical methods to solve these waveguide modes. For example, the rigorous vectorial coupled transverse mode integral equation formulation (VCTMIE), finite-difference frequency-domain method (FDFD), vectorial beam propagation method (VBPM) and modal expansion method with simple bases (MEMSB)¡Ketc. With the exception of the MEMSB method and the alike, all these methods work very hard to handle interface boundaries and many require many terms meet the convergence requirements. In this thesis, we propose a rigorous modal expansion method based on a set of orthogonal transverse magnetic bases to analyze arbitrarily profiled 2-D cylindrical dielectric waveguides. First, we expand the mode field solution of a general multi-layered waveguide by linear combination of 1D homogeneous solutions. Magnetic and components are chosen for its continuity property across the material interface. The choice of Magnetic field over electric field also reduces the number of terms and minimizes the Gibb¡¦s phenomenon. Our new vector bases eliminate the numerical difficulty of working with the singular term of the cylindrical differential operators. When compared with the results using simple bases, we further reduce one quarter of terms without loosing any accuracy. Although the process of deriving the formulation of this vector cylindrical basis expansion technique is complex because Bessel functions and their derivatives are involved, the resulting matrix eigenvalue-eigenvector equation is much simpler than that of the simple bases and the new the result is also more accurate. We also extended the analysis to study the 2-D cylindrical dielectric waveguide problem.

In this paper, we purpose a kernel-based fuzzy clustering algorithm to cluster data patterns in the feature space. Our method uses kernel functions to project data from the original space into a high dimensional feature space, and data are divided into groups though their similarities in the feature space with an incremental clustering approach. After clustering, data patterns of the same cluster in the feature space are then grouped with an arbitrarily shaped boundary in the original space. As a result, clusters with arbitrary shapes are discovered in the original space. Clustering, which can be taken as unsupervised classification, has also been utilized in resolving classification problems. So, we extend our method to process the classification problems. By working in the high dimensional feature space where the data are expected to more separable, we can discover the inner structure of the data distribution. Therefore, our method has the advantage of dealing with new incoming data pattern efficiently. The effectiveness of our method is demonstrated in the experiment.

Neural networks can be looked as a kind of intruments which is able to learn. For making the fruitful results of neural networks' learning possess parctical applied value, the thesis makes use of linear regression technics to strengthen the extended capability of RBF neural networks. The thesis researches the training methods of RBF neural networks, and retains the frame of OLS(orthogonal least square) learning rules which is published by Chen and Billings in 1992. Besides, aiming at the RBF's characteristics, the thesis brings up improved learning rules in first and second phases, and uses " early stop" to be the condition of training ceasing. To sum up, chiefly the thesis applies some technics of statistic linear regression to strenthen the extended capability of RBF, and using different methods to do computer simulation in different noise situations.

The lenses are important elements in optical imaging systems. However, lenses are liable to defects such as dusts and thus deteriorate their imaging quality. These kinds of imaging systems are investigated in this thesis .The polluted lens can be verified equivalent to a polluted random screen set against a clean lens .In our model ,the defects on random screen are assumed poisson-distribution ,overlapped and the transmittance effect of each defect is multiplicative .The autocorrelation function of screen is obtained by defects' density ,radius ,and transmittance. The evaluation of the optical transfer function for this imaging system can be achieved by the autocorrelation of the above random screen. This thesis includes computer simulation, experiments and comparison with other model and restoration method. The experiments are set up by the instruments including the video camera , capture card ,and personal computer. We may estimate the key parameters of our theoretical autocorrelation function by the real optical transfer function obtained from experiment. Accordingly, two methods are applied to image restoration in this thesis. One is to use the theoretical autocorrelation, the other is to use a second-order statistics of optical transfer function. The computation of second-order statistics involves a fourfold integration .By the help of changing variables and geometric analysis, we simplify the fourfold integration to double integration. Both of our methods are better for image restoration in RMS value than the method proposed by Tamas Daboczi

The shape of drift function in continuous time interest rate models has been investigated by many authors during the past decade. The main concerns have been whether the drift function is linear or nonlinear, but no convincing conclusions have been seen. In this dissertation, we investigate the reason for this problem and test several models of the drift function using a nonparametric test. Furthermore, we study some related problems, including the empirical properties of the nonparametric test. First, we propose regression models for the estimation of the drift function in some continuous time models. The limiting distribution of the parameter estimator in the proposed regression model is derived under certain conditions. Based on our analyses, we conclude that the effect of drift function for some U.S. Treasury Bill yields data is negligible. Therefore, neither linear nor nonlinear modeling has a significant effect. Second, parametric linear and nonlinear proposed regression models are applied and the correctness of those models is examined using the consistent nonparametric model specification test introduced by Li (1994) and Zheng (1996), henceforth the Jn test. The test results indicate that there is no strong statistical evidence against the assumed drift models. Furthermore, the constant drift model is not rejected either. Third, we compare the Jn and generalized likelihood ratio (GLR) tests through Monte Carlo simulation studies concerning whether the sizes of tests are stable over a range of bandwidth values, which is an important indicator to measure the usefulness of nonparametric tests. The GLR test was applied to testing the linear drift function in continuous time models by Fan and Zhang (2003). Our simulation study shows that the GLR test does not provide stable sizes over a grid of bandwidth values in testing the drift function of some continuous time models, whereas the Jn test usually does.

none