No description available.

This thesis describes the application of coupled-cluster theory to model systems of metallic solids and cold-atom gases. First, I give an overview of both ground- and excited-state coupled cluster theory as background for the main topics in this thesis. Next, I evaluate the accuracy of several cost-saving approaches in estimating the coupled cluster correlation energy for a model metallic system, the uniform electron gas, in the complete basis set and thermodynamic limits. After that, I present calculations of the spectral function of the uniform electron gas in these same limits, the results of which are rationalized by applying a bosonized coupled-cluster theory to an approximate, simplified Hamiltonian that couples plasmons to a structureless core hole state. Finally, I show how coupled-cluster theory captures the many-body nature of two-component Fermi gases with tunable, attractive interactions.

No description available.

Information fusion techniques have been widely applied in many applications including clustering, classification, detection and etc. The major objective is to improve the performance using information derived from multiple sources as compared to using information obtained from any of the sources individually. In our previous work, we demonstrated the performance improvement of Electroencephalography(EEG) based seizure detection using information fusion. In the detection problem, the optimal fusion rule is usually derived under the assumption that local decisions are conditionally independent given the hypotheses. However, due to the fact that local detectors observe the same phenomenon, it is highly possible that local decisions are correlated. To address the issue of correlation, we implement the fusion rule sub-optimally by first estimating the unknown parameters under one of the hypotheses and then using them as known parameters to estimate the rest of unknown parameters. In the aforementioned scenario, the hypotheses are uniquely defined, i.e., all local detectors follow the same labeling convention. However, in certain applications, the regions of interest (decisions, hypotheses, clusters and etc.) are not unique, i.e., may vary locally (from sources to sources). In this case, information fusion becomes more complicated. Historically, this problem was first observed in classification and clustering. In classification applications, the category information is pre-defined and training data is required. Therefore, a classification problem can be viewed as a detection problem by considering the pre-defined classes as the hypotheses in detection. However, information fusion in clustering applications is more difficult due to the lack of prior information and the correspondence problem caused by symbolic cluster labels. In the literature, information fusion in clustering problem is usually referred to as clustering ensemble problem. Most of the existing clustering ensemble methods are unsupervised. In this thesis, we proposed two semi-supervised clustering ensemble algorithms (SEA). Similar to existing ensemble methods, SEA consists of two major steps: the generation and fusion of base clusterings. Analogous to distributed detection, we propose a distributed clustering system which consists of a base clustering generator and a decision fusion center. The role of the base clustering generator is to generate multiple base clusterings for the given data set. The role of the decision fusion center is to combine all base clusterings into a single consensus clustering. Although training data is not required by conventional clustering algorithms (usually unsupervised), in many applications expert opinions are always available to label a small portion of data observations. These labels can be utilized as the guidance information in the fusion process. Therefore, we design two operational modes for the fusion center according to the absence or presence of the training data. In the unsupervised mode, any existing unsupervised clustering ensemble methods can be implemented as the fusion rule. In the semi-supervised mode, the proposed semi-supervised clustering ensemble methods can be implemented. In addition, a parallel distributed clustering system is also proposed to reduce the computational times of clustering high-volume data sets. Moreover, we also propose a new cluster detection algorithm based on SEA. It is implemented in the system to provide feedback information. When data observations from a new class (other than existing training classes) are detected, signal is sent out to request new training data or switching from the semi-supervised mode to the unsupervised mode. / Thesis / Doctor of Philosophy (PhD)

No / Software rejuvenation is a preventive and proactive fault management technique that is particularly useful for counteracting the phenomenon of software aging, aimed at cleaning up the system internal state to prevent the occurrence of future failure. The increasing interest in combing software rejuvenation with cluster systems has given rise to a prolific research activity in recent years. However, so far there have been few reports on the dependency between nodes in cluster systems when software rejuvenation is applied. This paper investigates the software rejuvenation policy for cluster computing systems with dependency between nodes, and reconstructs an stochastic reward net model of the software rejuvenation in such cluster systems. Simulation experiments and results reveal that the software rejuvenation strategy can decrease the failure rate and increase the availability of the cluster system. It also shows that the dependency between nodes affects software rejuvenation policy. Based on the theoretic analysis of the software rejuvenation model, a prototype is implemented on the Smart Platform cluster computing system. Performance measurement is carried out on this prototype, and experimental results reveal that software rejuvenation can effectively prevent systems from entering into disabled states, and thereby improving the ability of software fault-tolerance and the availability of cluster computing systems. / National Natural Science Foundation of China under the grant No. 60872044, 71133006, and Fundamental Research Funds for the Central Universities, and the Research Funds of Renmin University of China.

We give an overview of quantum chemical methods with a particular emphasis on the development of high-accuracy quantum chemical models. The reliability of these methods often hinges on whether enough electron correlation is included in the truncated wave function. As an example, we investigate the structures of m-benzyne and its fluorinated derivative, tetrafluoro-m-benzyne where the inclusion of triple excitations is paramount to correctly describe through-bond delocalization of the monocyclic form. At the CCSDT/6-31G** level of theory, the C1–C3 distance of the minimum energy form of m-benzyne is 2.0°A and the profile of the PES along the C1–C3 distance is that of an asymmetric, single-well, in agreement with previous density-functional theory and coupled cluster studies. In addition, the calculated CCSD(T) fundamental frequencies are in excellent agreement with the measured infrared frequencies, thus confirming the monocyclic form of m-benzyne. For tetrafluoro-m-benzyne, however, the increased eclipsing strain between the ring-external Câ X bonds stabilizes the bicyclo[3.1.0]hexatriene form: the C1–C3 distance is calculated at the CCSD(T)/cc-pVTZ level to be approximately 1.75 °A, which is in the range of elongated CC bonds. Computed harmonic vibrational frequencies compare reasonably well with the experimental neon-matrix difference spectrum and provide further evidence for the existence of a bicyclic form. We also report an extension of the coupled cluster iterative-triples model, CC3, to excited states of open-shell molecules, including radicals. We define the method for both spin-unrestricted Hartree-Fock (UHF) and spin-restricted open-shell Hartree-Fock (ROHF) reference determinants and discuss its efficient implementation in the PSI3 program package. The program is streamlined to use at most O(N7) computational steps and avoids storage of the triple-excitation amplitudes for both the ground-and excited-state calculations. The excitation-energy program makes use of a Lowdin projection formalism (comparable to that of earlier implementations) that allows computational reduction of the Davidson algorithm to only the single- and double-excitation space, but limits the calculation to only one excited state at a time. However, a root-following algorithm may be used to compute energies for multiple states of the same symmetry. Benchmark applications of the new methods to the lowest valence 2B1 state of the allyl radical, low-lying states of the CH and CO+ diatomics, and the nitromethyl radical show substantial improvement over ROHF- and UHF-based CCSD excitation energies for states with strong double-excitation character or cases suffering from significant spin contamination. For the allyl radical, CC3 adiabatic excitation energies differ from experiment by less than 0.02 eV, while for the 2§+ state of CH, significant errors of more than 0.4 eV remain. Finally, ground- and excited-state dipole moments are derived diagramatically and were recently developed within the PSI3 quantum chemistry package. However, convergence problems with computing the left-hand excited-state has prevented us from reporting any meaningful results. Thus, future work includes solving this convergence problem before the effects of triple excitations on one-electron properties can be reported with certainty. / Ph. D.

AgaR, a protein in <i>Escherichia coli</i> thought to control the metabolism of N-acetylgalactosamine, is a member of the DeoR family of transcriptional regulators. Three transcriptional promoters within a cluster of genes containing the gene for AgaR were identified, specific for <i>agaR, agaZ</i> and <i>agaS</i>, and the transcription start sites mapped. Transcription from these promoters was specifically induced by N-acetylgalactosamine or galactosamine, though K-12 strains lacked the ability to utilize these as sole sources of carbon. The activity of these promoters was constitutively elevated in a strain in which <i>agaR</i> had been disrupted confirming that the promoters are subject to negative regulation by AgaR. AgaR-His6, purified using immobilized metal affinity chromatography, was used for DNase I footprint analysis of the promoter regions. Four operator sites bound by AgaR were identified. A putative consensus binding sequence for AgaR was proposed based on these four sites. <i>In vivo</i> and <i>in vitro</i> analysis of the <i>agaZ</i> promoter indicated that this promoter was activated by the cAMP-cAMP receptor protein (CRP). Expression from the <i>aga</i> promoters was less sensitive to catabolite repression in revertants capable of <i>N</i>-acetylgalactosamine utilization, suggesting that these revertants have mutation(s) that result in an elevated level of inducer for AgaR. A cluster of genes at minute 87.7 of the <i>E. coli</i> genome contains a gene that encodes another member of the DeoR family of transcriptional regulators. This protein, YihW, is more similar to GlpR, transcriptional regulator of <i>sn</i>-glycerol 3-phosphate metabolism in <i>E. coli</i>, than other members of the DeoR family. Despite the high degree of similarity, YihW lacked the ability to repress P_glpK, a promoter known to be controlled by GlpR. A variant of YihW containing substitutions in the putative recognition helix to more closely match the recognition helix of GlpR was also unable to repress P_glpK. Transcriptional promoters identified in this cluster of genes were negatively regulated by YihW. Regulation of genes involved in the metabolism of <i>sn</i>-glycerol 3-phosphate in <i>E. coli</i> by GlpR has been well characterized. However, the function of a protein (GlpE) encoded by a gene cotranscribed with that for GlpR was unknown prior to this work. GlpE was identified as a single-domain, 12-kDa rhodanese (thiosulfate:cyanide sulfurtransferase). The enzyme was purified to near homogeneity and characterized. As shown for other characterized rhodaneses, kinetic analysis revealed that catalysis occurs via an enzyme-sulfur intermediate utilizing a double-displacement mechanism requiring an active-site cysteine. K_m (SSO₃²⁻) and K_m (CN⁻) were determined to be 78 mM and 17 mM, respectively. The native molecular mass of GlpE was 22.5 kDa indicating that GlpE functions as a dimer. GlpE exhibited a kcat of 230 s-1. Thioredoxin, a small multifunctional dithiol protein, served as sulfur-acceptor substrate for GlpE with an apparent K_m of 34 mM when thiosulfate was near its K_m, suggesting thioredoxin may be a physiological substrate. / Ph. D.

A system was developed for the controlled expression of genes in Azotobacter vinelandii by using genomic fusions to the sucrose catabolic regulon. This system was used for the functional analysis of the A. vinelandii isc genes, whose products are involved in the maturation of [Fe-S] proteins. For this analysis the scrX gene, contained within the sucrose catabolic regulon, was replaced by the A. vinelandii iscS, iscU, iscA, hscB, hscA, fdx, iscX gene cluster, resulting in duplicate genomic copies of these genes, one whose expression is directed by the normal isc regulatory elements (Pisc) and the other whose expression is directed by the scrX promoter (PscrX). Functional analysis of [Fe-S] protein maturation components was achieved by placing a mutation within a particular Pisc-controlled gene with subsequent repression of the corresponding PscrX-controlled component by growth on glucose as the carbon source. This experimental strategy was used to show that IscS, IscU, HscBA and Fdx are essential in A. vinelandii and that their depletion results in a deficiency in the maturation of aconitase, an enzyme that requires a [4Fe-4S] cluster for its catalytic activity. Depletion of IscA results in null growth only when cells are cultured under conditions of elevated oxygen, marking the first null phenotype associated with the loss of a bacterial IscA-type protein. Furthermore, the null growth phenotype of cells depleted for HscBA could be partially reversed by culturing cells under conditions of low oxygen. These results are interpreted to indicate that HscBA and IscA could have functions related to the protection or repair of the primary IscS/IscU machinery when grown under aerobic conditions. Conserved amino acid residues within IscS, IscU, and IscA that are essential for their respective functions and/or display a partial or complete dominant-negative growth phenotype were also identified using this system. Inactivation of the IscR repressor protein resulted in a slow growth phenotype that could be specifically attributed to the elevated expression of an intact [Fe-S] cluster biosynthetic system. This system was also used to investigate the extent to which the two [Fe-S] biosynthetic systems in A. vinelandii, Nif and Isc, can perform overlapping functions. Under normal laboratory growth conditions, no cross-talk between the two systems could be detected. However, elevated expression of Isc components as a consequence of inactivation of the IscR repressor protein results in a modest ability of the Isc [Fe-S] protein maturation components to replace the function of Nif-specific [Fe-S] protein maturation components. Similarly, when expressed at very high levels the Nif-specific [Fe-S] protein maturation components could functionally replace the Isc components. Oxygen levels were also found to affect the ability of the Nif and Isc systems to perform common functions. Nevertheless, the lack of significant reciprocal cross-talk between the Nif and Isc systems when they are produced only at levels necessary to satisfy their respective physiological functions, indicates a high level of target specificity with respect to [Fe-S] protein maturation. / Ph. D.

One of the most promising methods for surmounting the high-degree polynomial scaling wall associated with electron correlating wave function methods is the local correlation technique of Pulay and Saeb&oslash;. They have proposed using a set of localized occupied and virtual orbitals free of the canonical constraint commonly employed in quantum chemistry, resulting in a method that scales linearly (in the asymptotic limit) with molecular size. Pulay and Saeb$oslash; first applied their methods to configuration interaction and later to M$oslash;ller-Plesset perturbation theory. Werner et al. have have extended the local correlation scheme of Pulay and Saeb$oslash; to coupled-cluster theory. One of the pitfalls of the local correlation methods developed by Pulay and Saeb$oslash; is the dependence of domain selection on the molecular geometry. In other words, as the geometry changes the domain structure of the local correlation calculation can change also, leading to discontinuities in the potential energy surface. We have examined the size of these discontinuities for the homolytic bond cleavage of fluoromethane and the heterolytic bond dissociation of singlet ketene and propadienone. Properties such as polarizabilities and optical rotation are realized through linear response theory, where the Hamiltonian is subject to an external perturbation and the wave function is allowed to respond to the applied perturbation. Within the context of local correlation it is necessary to understand how the domain structure alters in response to an applied perturbation. We have proposed using solutions to the CPHF equations (coupled-perturbed Hartree-Fock) in order to predict the correlation response to an applied perturbation. We have applied this technique to the calculation of polarizabilities, with very favorable results, and also to optical rotation, with mixed results. / Ph. D.

Coupled cluster (CC) and density functional theory (DFT) are highly regarded as robust quantum chemical methods for accurately predicting a wide variety of properties, such as molecular structures, thermochemical data, vibrational spectra, etc., but there has been little focus on the theoretical prediction of optical rotation. This property, also referred to as circular birefringence, is inherent to all chiral molecules and occurs because such samples exhibit different refractive indices for left- and right- circularly polarized light. This thesis focuses on the theoretical prediction of this chiroptic property using CC and DFT quantum chemical models. Several small chiral systems have been studied, including (S)-methyloxirane, (R)-epichlorohydrin, (R)-methylthiirane, and the conformationally flexible molecules, (R)-3-chloro-1-butene and (R)-2-chlorobutane. All predicted results have been compared to recently published gas-phase cavity ringdown polarimetry data. When applicable, well-converged Gibbs free energy differences among confomers were determined using complete-basis-set extrapolations of CC energies in order to obtain Boltzmann-averaged specific rotations. The overall results indicate that the theoretical rotation is highly dependent on the choice of optimized geometry and basis set (diffuse functions are shown to be extremely important), and that there is a large difference between the CC and DFT predicted values, with DFT usually predicting magnitudes that are larger than those of coupled cluster theory. / Ph. D.