Search

We are interested in the efficient numerical solution of the structured linear systems that arise when we apply stochastic Galerkin mixed finite element methods (SGMFEMs) to systems of partial differential equations (PDEs) with inputs that depend on a possibly large number of uncertain parameters. In particular, we are interested in PDEs arising in engineering applications for which SGMFEMs give rise to saddle point systems, such as linear elasticity and fluid flow problems. Despite being highly structured, saddle point systems associated with SGMFEMs are challenging to solve due to their extremely large size. The number of equations is the product of (i) the number of degrees of freedom associated with the chosen mixed finite element method on the spatial domain and (ii) the dimension of the polynomial space associated with the parameter domain. When we refine the finite element mesh and/or increase the degree of the parametric polynomial approximation to improve accuracy, the dimension of the associated linear system increases. When working on standard desktop computers, one cannot use conventional Krylov subspace methods for very fine SGMFEM discretisations because storing the required matrices and vectors quickly exhausts available memory. One potential remedy is to recast the discrete problem as a linear multi-term matrix equation (LMTME) and use reduced basis methods. Such methods construct low rank factored approximations to the solution matrix by projecting the problem onto a lower-dimensional space. Our main aim is to develop a memory efficient solver for the discrete problems that arise when we apply SGMFEMs to a three-field linear elasticity model with parameter-dependent Young modulus. The starting point is a reduced basis method known as Multi-RB that was recently proposed for LMTMEs associated with symmetric and positive definite problems. After reformulating the matrix equation, the scheme iteratively constructs a reduced basis using a strategy inspired by rational Krylov subspace approximation, then applies Galerkin projection and solves a reduced problem. When the problem is not positive definite, the best choice of preconditioner and projection technique to apply is not clear. For large-scale problems with solutions that cannot be approximated well by very low rank matrices, selecting a computationally feasible stopping condition is also problematic. We modify the Multi-RB method to solve LMTMEs associated with symmetric and indefinite problems. For the linear elasticity problem of main interest, we critically assess the performance of the method using two distinct preconditioning strategies and two projection techniques. We also provide new eigenvalue analysis for the preconditioned systems. For smaller problems, we examine the convergence of the method by tracking the preconditioned and unpreconditioned residuals. Finally, to compare how the solver behaves on a different LMTME with the same structure, but whose solution matrix has different rank characteristics, we also consider a two-field groundwater flow model with parameter-dependent permeability coefficient.

Timber has become one of the popular choices as a sustainable building material, and with the increase in timber building height, the problem caused by wind load becomes remarkable. Most research focused on the ultimate structure performance under wind load or vibration under cyclic shear force due to dynamic wind load. However, a lack of research on the serviceability of timber wall connections under out-of-panel load caused by dynamic wind pressure may lead to underestimating damage to timber structures. Therefore, this study investigates the fatigue performance of connections for plug-in timber panels under out-of-panel cyclic load caused by wind. The study first developed a new method to design loading sequences for a lab test, to analyse the joint fatigue behaviour under wind load. This research focused on the angle bracket connection system with fasteners. Compared with the testing sequences from other building codes, the newly developed process is more logical and can save testing time. The high-cycle fatigue testing results proved that wind load could degrade stiffness of timber wall joint with an angle bracket system but did not cause the connection system failure or material damage. Additionally, with the increase in building height, stiffness reduction increased. The angle bracket connecting system can connect timber and concrete but selecting the angle bracket type is essential. Different connecting details were tested. The results demonstrate that changing nail pattern or angle bracket size can influence the cyclic fatigue performance of timber wall joints. By applying the digital image correlation (DIC), this study indicates that the cyclic loading mainly affects the mechanical properties of joints by affecting the embedment of fasteners rather than the strain of the angle bracket. This study then extended the investigation to the life-cycle environmental impact of timber building. Results pointed out that neglecting the energy or material used for joint timber maintenance can cause at least 20% underestimating the embodied carbon.

Online learning has been reported to be increasingly used within Chinese higher education. The rapid expansion has created a problem that it's difficult to carry out real-time learner interaction among a great number of students. The pure online instruction made the existing problems in university EFL teaching even worse, such as 'teacher-centred' instruction, student's low learning motivation and low level of class participation. There is a need to develop pedagogies in EFL teaching in China to meet expanding demand and solve the problems. Therefore, this study aimed to explore the integration of cooperative learning (CL) into online EFL curriculum to facilitate learner interactions, bring about meaningful learning together and research its influence on learner writing proficiency. To be specific, it was to explore whether online cooperative learning (OCL) with elements of Social Interdependence Theory (SIT) could actually manifest during learning and teaching online and whether there may be initial evidence that OCL could be effective at raising EFL learners' writing attainment. A researcher-designed online writing curriculum was delivered to 17 university students in nine weeks, once per week for 90 minutes each session. The course, named Step by Step University Writing Skills, was composed of 12 topics, aiming to develop learner's skills of writing a sentence, writing a paragraph and writing an argumentative essay step by step. Learning tasks using CL techniques Jigsaw, Think-Pair-Share (TPS), Random Reporter, Assigning group roles (AGR), were assigned in each session for the learners to practice the teaching contents in groups. The Questionnaire-Scale of 'Learner Experience & Perception on Cooperative Learning' was used to assess the learners' prior knowledge of CL, their previous experience of CL and their attitudes toward CL before they took the OCL course. Learners' writing proficiency tests were taken twice before and after the course as the pre-test and the post-test to measure the potential impact of OCL pedagogy iii (notwithstanding that this was a modelling trial and did not use a control group). Learners' online cooperative interactions were video-recorded, transcribed and analyzed qualitatively to investigate the evidence of SIT in CL online. A post-course questionnaire was given after the intervention to find out learners' perceptions of the impact of CL implemented in the study. As this was a small modelling trial, the study was not powered to detect the effectiveness of the intervention. Rather, it was to explore the feasibility of CL pedagogies be used in online curriculum for EFL learners. The analysis showed that CL with elements of SIT can be identified in learner online interactions. Learners' writing attainment was improved after the course. The conclusion was that CL activities could be designed into an online course with EFL students; students reported that using the OCL pedagogies was an acceptable way to learn; that the technique could be scaled up to a larger sample. The study provided evidence on how OCL worked and how the online pedagogy was structured for CL to work. It presented a good theoretical model for planning and structuring CL activities online with EFL students. The study offered a basis to conduct a larger effectiveness study of the OCL. The findings in this study contributed to the existing literature on online cooperative language learning and CL in EFL instruction. Besides, it combined the cooperative skills into the curriculum contents, which contributed greatly to the higher education practice from the social dimension. In doing so, it offered an opportunity to make a positive contribution to the higher education reform in China.

This thesis examines how different types of human-dog relationship are constructed and how they interplay dynamically in Yulin, where the widely criticised dog-meat festival is held each year. Through an 11-month period of multispecies ethnographic fieldwork, I found that dogs were used for guarding, hunting, fighting, companionship, and food. The first three types of relationship are part of traditional rural life, while the practice of keeping companion dogs as pets is a Western-oriented urban practice, developed through cultural globalisation. The rural relationships are in decline, while the latter two are expanding with urbanisation. Dog-meat eating, originally an occasional rural practice, has transformed into an urban business form, with an economic chain. Broader social processes are fused into the ways that people from different social backgrounds 'do gender'; these include the various types of relationship people construct with dogs. Dogs with good working performance are constructed with rural masculinities identified by rural men, while dogs kept as pets are perceived as cute, constructing femininities for urban young women. Meanwhile, human-dog power dynamics vary with gender, as rural men's affection for working dogs serves their dominance over dogs, while urban woman's affection results in an ethics of care for dogs, counterbalancing this compensated power. Under this compensated power relationship, many women refrain from eating dog meat. However, their voices are marginalised by local men's strong defence of dog meat, which is believed to strengthen men's health. Amidst controversy about the festival, a discourse that categorises dogs into 'pet dogs' and 'Tugo' has become widespread in Yulin. 'Tugou' is the indigenous type of dog used traditionally for guarding, hunting, fighting, and as food. They are generally much less valued than pet dogs. These different social constructions of dogs result in 'dog racism' in Yulin.

This thesis explores the BBC programme, Question Time, treating it as an instantiation of Audience Participation Political Debate (APPD). It examines the nature of political conversation with a focus on the relationship between the practices of overlap and intersubjectivity. This study also addresses the question of why overlap occurs frequently during the course of the conversation in Question Time and how it contributes to the construction, restoration and maintenance of intersubjectivity. The data consists of 16-hours of recordings of Question Time that were aired between January and June 2016. This study uses Conversation Analysis to uncover and examine the sequence organization of overlapping talk in terms of overlapping questions, overlapping statements and overlapping applause. The findings demonstrate that in this APPD context overlapping questions function to solicit the breakdown of intersubjectivity and project the next turns to do repairs. In contrast, overlapping statements and applause serve as an indication of agreement and disagreement among participants, which are found to be the sites on which intersubjectivity rests. The findings of this study contribute to a better understanding of APPD and how in this genre of hybrid media formats the practice of overlap functions. This media format provides an arena that allows the participants greater opportunities to defend and repair their political stances, as well as to display agreements and negotiate disagreements. In addition, the findings add new knowledge to our understanding of the architecture of intersubjectivity in APPD that intersubjectivity can be constructed via the mechanism of negotiation and restored and maintained via repair mechanism.

This thesis critiques the concept of narrative coherence through an in-depth inquiry into the lived experience of narrative incoherence in counselling and psychotherapy. It questions whether lived experience necessarily requires narrative structure and the extent to which a coherent narrative is essential for psychological and emotional well-being. It thus attempts to depathologise the experience of narrative incoherence, instead honouring those moments in therapy when words quite literally fail us. Adopting writing as inquiry as its methodological foundation, the thesis continuously moves between experiential self-searching and intellectual engagement with theoretical insights. I draw particularly on conceptual resources offered by Butler, Foucault, Wittgenstein, Winnicott, Freud, Leader and the theory of the dialogical self. I write into my personal experiences as a client in therapy and as a therapist and this analytical work is complemented by reflections on my experience of a short-term sandplay process, undertaken specifically for this research project, which aimed to surface the interstices of language and the non-verbal therapeutic process. Parallel to its questioning of the demand for narrative coherence is this thesis's challenge to the linear, well-structured and well-articulated scholarly voice of traditional doctoral thesis work. Methodologically and stylistically, this thesis stays close to that which is inarticulate and unstructured, which is often termed incoherence. Instead of presenting a planned linear process, I argue that the research process of this thesis is an unforeseeable and unpredictable journey into the unknown in which I encounter rather than choose the conceptual resources I use. This thesis concludes with three main points, namely 'letting go', 'holding on' and 'to play'. I encourage therapists to let go of the obsession with and demand for coherent narrative as these can silence, alienate and pathologise individuals. Meanwhile, the research process also highlights the necessity of telling one's story to a caring other, and the predicament between the impossibility of expression and the necessity to tell and to be known. In relation to this, I encourage therapists to stay with and listen to what is unsayable and unnarratable without demand. Lastly, this thesis puts forth the value of playfulness when working with the concept of narrative coherence and incoherence, calling for a blurring boundary between the two.

Manufacturing enterprises nowadays face the challenge of increasing energy prices and requirements to reduce their emissions. Most reported work on reducing manufacturing energy consumption focuses on the need to improve the efficiency of resources (machines). The potential for energy reducing at the system-level has been largely ignored. At this level, operational research methods can be employed as the energy saving approach. The advantage is clearly that the scheduling and planning approach can be applied across existing legacy systems and does not require a large investment. For the emission reduction purpose, some electricity usage control policies and tariffs (EPTs) have been promulgated by many governments. The Rolling Blackout policy in China is one of the typical EPTs, which means the government electricity will be cut off several days in every week for a specific manufacturing enterprise. The application of the Rolling Blackout policy results in increasing the manufacturing enterprises’ costs since they choose to start to use much more expensive private electricity to maintain their production. Therefore, this thesis develops operational research methods for the minimisation of electricity consumption and the electricity cost of job shop type of manufacturing systems. The job shop is selected as the research environment for the following reasons. From the academic perspective, energy consumption and energy cost reduction have not been well investigated in the multi-objective scheduling approaches to a typical job shop type of manufacturing system. Most of the current energy-conscious scheduling research is focused on single machine, parallel machine and flow shop environments. From the practical perspective, job shops are widely used in the manufacturing industry, especially in the small and medium enterprises (SMEs). Thus, the innovative electricity-conscious scheduling techniques delivered in this research can provide for plant managers a new way to achieve cost reduction. In this thesis, mathematical models are proposed for two multi-objective job shop scheduling optimisation problems. One of the problems is a bi-objective problem with one objective to minimise the total electricity consumption and the other to minimise the total weighted tardiness (the ECT problem). The other problem is a tri-objective problem which considers reducing total electricity consumption, total electricity cost and total weighted tardiness in a job shop when the Rolling Blackout policy is applied (the EC2T problem). Meta-heuristics are developed to approximate the Pareto front for ECT job shop scheduling problem including NSGA-II and a new Multi-objective Genetic Algorithm (GAEJP) based on the NSGA-II. A new heuristic is proposed to adjust scheduling plans when the Rolling Blackout policy is applied, and to help to understand how the policy will influence the performance of existing scheduling plans. NSGA-II is applied to solve the EC2T problem. Six scenarios have been proposed to prove the effectiveness of the aforementioned algorithms. The performance of all the aforementioned heuristics have been tested on Fisher and Thompson 10×10, Lawrence 15×10, 20×10 and 15×15 job shop scenarios which were extended to incorporate electrical consumption profiles for the machine tools. Based on the tests and comparison experiments, it has been found that by applying NSGA-II, the total non-processing electricity consumption in a job shop can decrease considerably at the expense of the schedules’ performance on the total weighted tardiness objective when there are tight due dates for jobs. When the due dates become less tight, the sacrifice of the total weighted tardiness becomes much smaller. By comparing the Pareto fronts obtained by GAEJP and by NSGA-II, it can be observed that GAEJP is more effective in reducing the total non-processing electricity consumption than NSGA-II, while not necessarily sacrificing its performance on total weighted tardiness. Thus, the superiority of the GAEJP in solving the ECT problem has been demonstrated. The scheduling plan adjustment heuristic has been proved to be effective in reducing the total weighted tardiness when the Rolling Blackout policy is applied. Finally, NSGA-II is proved to be effective to generate compromised scheduling plans for using the private electricity. This can help to realise the trade-off between the total weighted tardiness and the total electricity cost. Finally, the effectiveness of GAJEP in reducing the total non-processing electricity consumption has been validated in a real-world job shop case.

Fire is one of the main hazards associated with storage tanks containing flammable liquids. These tanks are usually closely spaced and in large groups, so where a petroleum fire occurs, adjacent tanks are susceptible to damage leading to further development of the fire. The structural behaviour such as thermal stability and failure modes of the tanks under such fire scenario are very important to the safety design and assessment of oil depots. However, no previous studies on this problem are known to the best knowledge of the author. This thesis presents a systematic exploration of the potential thermal and structural behaviours of an oil tank when one of its neighbour tanks is on fire. Under such scenario, the oil tanks are found to easily buckle under rather moderate temperature rises. The causes of such buckling failures are the reduced modulus of steel at elevated temperatures, coupled with thermally-induced stresses due to the restraint of thermal expansion. Since the temperatures reached in such structures can be several hundred Centigrade degrees, any restraint to thermal expansion can lead to the development of compressive stresses. The high susceptibility of thin shell structures to elastic buckling under low compressive stresses means that this type of failure can be easily provoked. The main objectives of this thesis were to reveal the thermal distribution patterns developed in an oil tank under the heating from an adjacent tank fire, to understand the underlying mechanism responsible for the buckling of tank structure, and to explore the influences of various thermal and geometrical parameters on the buckling temperature of the tanks. The study began with analytical solutions for stresses and deformations in a partially filled roofless cylindrical tank under an idealised axisymmetrical heating regime involving thermal discontinuity at the liquid level. The results demonstrate that large compressive circumferential membrane stresses occur near the bottom boundary for an empty tank and near the liquid level for a partially-filled tank. Heat transfer analysis was conducted to explore the temperature distribution developed in the tank when the fire reaches a steady state. Parameters and assumptions used in the adopted pool fire model were carefully examined. The results show that a rather non-uniform distribution of temperature is developed in the tank especially around the tank circumference. A simple model was then proposed to describe the temperature distribution based on the numerical heat transfer analysis. The accuracy of the proposed temperature distribution model for predicting the structure behaviour was evaluated by comparing its predictions with those using directly the temperature distribution obtained from the numerical heat transfer analysis. Extensive geometric and material nonlinear analyses were carried out to capture the buckling behaviour of the tank using both the proposed temperature distribution and that from heat transfer analysis. It was found large vertical compressive membrane stresses are induced in the tank, causing buckling. The influence of fire diameter, location, liquid filling level and tank geometry were investigated.

The thesis presents study to explore the role of dynamic features in speaker verification. Based on the theory that dynamic information should contain important speaker information, modelling the dynamics should have the potential to improve the speaker verification performance. Experiments on TD-SV using segmental hidden Markov models (SHMMs) on the YOHO database show performance improvement. However there is no significant improvement for TI-SV from experiments on the Switchboard database, using segmental GMMs. Analysis of the TD-SV results confirms that the speech dynamics modeled by SHMMs contribute more to the SV accuracy. Analysis of the TI-SV results indicates that the lack of speech dynamic information is a feature of GMM systems. It seems that the priority of the maximum likelihood training algorithm is to model stationary regions, and the role of dynamic features in GMM system, is to ensure that the classification focuses on static regions rather than to model dynamics. Study on TI-SV was carried out using conventional GMMs. Without RASTA filtering, the `delta-only' system works best. However, after RASTA filtering, the `static-plus-delta' system performs best. The results suggest that the good performance of the `delta-only' system before RASTA is mainly due to the noise robustness of the delta parameters.

It is essential to always track maximum power from photovoltaic (PV) sources. Failure to track the global maximum power point under partial shading conditions is one of the major reasons that lead to significant power losses. Several maximum power point tracking methods have been proposed to deal with this problem. However, none of them were able to effectively identify the occurrence of partial shading. With the facility of Matlab modelling and simulation as well as the aid of a constructed solar emulator, the power-voltage characteristics of a PV panel under uniform and non-uniform irradiance conditions have been studied and some useful conclusions have been identified from observation. Based on these conclusions, a novel maximum power point tracking algorithm has been proposed, which is capable of identifying the occurrence of partial shading hence determining the need for a global scan over the operation range of PV panels for the true maximum power point. In the meantime, the effect of PV dynamics, due to the capacitance of PV cells, on maximum power point trackers has been investigated and some initial results and suggestions have been presented in this work.

Testing an evolutionary framework, this study examined moral self-enhancement in relation to self-deception and self-construal in a cross-cultural context. The participants included 127 U.S. and 107 Chinese college students. The results demonstrated that moral self-enhancement is not a characteristic unique to individualistic ideology but rather a universal motivation. Regardless of their cultural groups and self-construal, participants tended to morally self-enhance, rating their own character and sense of responsibility significantly higher than those of other people. In addition, U.S. participants were more likely to morally self-enhance compared to their Chinese counterparts. At the individual level, strong independents demonstrated greater moral self-enhancement than did their strong interdependent peers among the U.S. participants, whereas greater moral self-enhancement was observed in strong interdependents but not strong independent in the Chinese sample. As hypothesized, self-deception as measured by "self-deceptive enhancement" (SDE) but not "impression management" (IM) stood out as a significant predictor of moral self-enhancement, supporting the evolutionary understanding that moral self-enhancement is an unconscious process intimately related to self-deception. Among Chinese participants, the association between moral self-enhancement and self-deception was found to be mediated through an inflated rating of others, which indicated a potential other-enhancement effect. On the one hand, results of this study regarding cultural differences in the demonstration of moral self-enhancement were in line with extant self-enhancement literature. On the other hand, the significant relationship between moral self-enhancement and self-deception as revealed in the study provided evidence against the claim that culture is the primary explanation for self-enhancement. Moral self-enhancement as a psychological adaptation to cooperation in the social environment has an evolutionary root. Findings of the study also suggested challenges facing current moral education practices. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

This study is one of only a few that attempts to examine simultaneously faculty productivity in teaching, research, and service. The research is guided by a conceptual model built from several branches of the literature on career stage theory, motivation theory, and previous studies of faculty productivity. The model hypothesizes that faculty engagement in different scholarly activities is shaped by the combined influence of personal and contextual factors and that the effects of those factors vary over a faculty member's career to produce different activity profiles across the areas of faculty effort. The model guides the research design, selection of variables, and statistical methods. Data for the study are obtained from the Higher Education Research Institute (HERI) and consist of a population of over 3,400 Science, Technology, Engineering, and Math (STEM) faculty members in research and doctoral-granting institutions who responded to the 2004 HERI Faculty Survey. Using Principle Axis factor analysis, the survey responses are reduced to clusters of variables reflecting nearly all the concepts in the model. The study also develops multiple measures of each type of faculty productivity. This study extends the scope of previous research by using multiple measures of each type of academic productivity and Latent Profile Analysis to identify profiles of faculty productivity. The results of this study provide insights into faculty productivity in several important ways. First, the study finds empirical support for a conceptual model linking faculty instructional, research, and service productivity in a single conceptual framework. Based on this model, this study extends the scope of previous scholarship by using multiple measures of the three different types of academic activity, and by examining the interaction of personal and organizational factors on faculty productivity using both simple multinomial logistic regression and Hierarchical General Linear Modeling. Second, the LPA results are quite robust and indicate that the 13 measures of faculty productivity used in the study identify four, very clear categories: faculty members who are primarily research oriented, primarily teaching oriented, primarily oriented toward internal service (such as campus administration and governance), and primarily oriented toward external outreach. Consistent with the mission of universities, the largest group (57.3%) are research-oriented faculty, followed by teaching-oriented faculty (31.9%), internal service oriented (5.6%), and external outreach (5.2%). The LPA results produced an unusually clean solution with the average probabilities for membership in each of the four classes ranging from 79 to 89 percent of the cases correctly classified. Third, although causality is not clear, the model-driven results confirm that faculty behavior may be influenced by a combination of many individual and institutional factors. As suggested by career stage theories and motivation theories, career age and intrinsic motivation, like research interests and needs for self-actualization are the best individual level predictors of faculty productivity patterns. Organizational contexts like institutional control, mission emphasis, admission selectivity and campus climate for research and faculty administration relations also exert strong influences on faculty work preference, and steer faculty towards certain production functions. Fourth, the current study contributes to our understanding of administrative practice as an influence on faculty work, and may have important implications for reward systems and faculty development. The research results here indicate that few faculty members attain above-average productivity simultaneously in teaching, research, and service. Instead, they appear to concentrate their efforts on particular areas of activity. Thus, a more realistic institution policy might be to assess and encourage balance among these different professional activities at the aggregated department level rather than at the individual level. Although this study is not longitudinal, the results suggest that faculty members shift their attention to certain production functions at distinct periods in their careers. Therefore, instead of expecting faculty members to be equally productive in teaching, research and service at the same time, higher education institutions might better differentiate individual faculty responsibilities and provide (even promote) diverse career advancement paths for faculty members with variable interests in teaching, research and service. Finally, the current findings suggest that faculty development may play an important role in engaging faculty members, particularly in teaching activities. Extrinsic rewards, such as salaries, appear to have limited impact on faculty behavior compared to the internal needs, values, and interests faculty bring with them to the job. To improve instructional effectiveness, institutions should alter faculty values and enhance faculty commitment towards teaching via carefully crafted faculty development and mentoring programs, instead of relying solely on increasing extrinsic rewards for teaching. Moreover, efforts to improve faculty productivity in any given area may need to begin at the front end, by recruiting faculty with known preferences, rather than by attempting to change them after hiring. (Abstract shortened by UMI.). [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

Tables are ubiquitous with a history that pre-dates that of sentential text. Authors often report a summary of their most important findings using tabular structure in documents. For example, scientists widely use tables to present the latest experimental results or statistical data in a condensed fashion. Along with the explosive development of the digital library and Internet, tables have become a valuable information source for information seeking and data analysis. Interest in and use of table data necessitates table indexing and search. However, current search engines do not support table search. The difficulty of automatically extracting tables from un-tagged documents, the lack of a universal table metadata specification, and the limitation of the existing ranking schemes make the table search problem challenging. Effectively and efficiently searching table data becomes an urgent demand. In this dissertation, we present an automatic table extraction and search engine, "TableSeer". "TableSeer" crawls the web and digital libraries, detects tables from documents using heuristic-based and machine-learning based methods, represents tables using an extensive set of medium-independent table metadata that other people can reuse, indexes table metadata files, ranks tables, and provides a user-friendly search interface. To improve the performance of the table boundary detection, a novel page-box-cutting method and a sparse-line detection method are proposed. Given a keyword-based table search query, TableSeer ranks the matched tables and returns the most relevant tables with a novel table ranking algorithm--TableRank. TableRank tailors the classic vector space model and adopts an innovative term weighting scheme by aggregating multiple features from three levels: the term, table and document levels. Although tables are widely used, there is no standard on the table structure designing. Many issues that go into the design of tables and will impair the table data readability, accessibility, and reusability are ignored. In order to have a deep understanding on the table characterization and to improve the table extraction and search performance, we also implement the first large-scale table quantitative study on table natures in digital libraries. We demonstrate the value of "TableSeer" with empirical studies on scientific documents. The experimental results show that our table search engine outperforms existing search engines on table search. Overall, TableSeer eliminates the burden of manually extracting table data from digital libraries and enables users to automatically examine tables. TableSeer is successfully deployed and in current use in several scientific digital libraries, for example "CiteSeer[superscript x]". [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

The graphic rating scale, a measurement tool used in many areas of psychology, usually takes a form of a fixed-length line segment, with both ends bounded and labeled as extreme responses. The raters mark somewhere on the line, and the length of the line segment from one endpoint to the mark is taken as the measure. An item response unfolding model is proposed to analyze the bounded continuous data collected with the graphic rating scale. The model has both location and dispersion parameters, and the item response function is developed based on the truncated normal density. The item parameters are estimated using maximum marginal likelihood estimation, and the standard errors of the estimates are computed from the observed information matrix. Simulation studies were conducted to investigate the behavior of the estimators in two simple versions of the model, one with only location parameters and the other with a common dispersion parameter. Survey data from the American National Election Study were used to demonstrate the application of the proposed model in studying people's opinions towards political figures or social groups in the 2008 presidential election. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

This thesis presents results for two polymeric systems with liquid-to-solid transitions, i.e., oppositely charged polyelectrolytes in water with salt that can form a gel and polydimethylsiloxane (PDMS)-particle suspensions that exhibit yield stress in 3D printing applications. A protocol is developed for quantifying the compositions of both supernatant and coacervate phases using a combination of H NMR, C NMR, TGA, conductivity, and titration. With these methods, the phase diagrams of polyacrylic acid /poly (diallyl dimethylammonium chloride) (i.e., PAA/PDADMAC) in potassium chloride at different pH values are generated experimentally. We find that with increasing salt concentration, the coacervate phase volume initially increases and then quickly drops to zero at the critical salt concentration, and we quantify the concentrations of both polyelectrolytes in both phases for some compositions. At low pH, we also find a novel phase separation re-entry at high salt concentration, probably related to the solubility of PAA as a function of pH and salt, implying strong non-electrostatic driving forces for coacervation at low pH. The linear viscoelasticity of coacervates is reviewed, with a focus on time-temperature/salt/pH/hydration superpositions. A variety of polyelectrolyte pairs show successful time-salt superposition, with master curves similar to those for neutral polymers. However, in some cases, a solid-like, as opposed to a fluid-like, response is observed at low frequencies, especially at low salt concentrations. Some coacervates seem to fit the “sticky diffusion” theory reasonably well, wherein relaxation is controlled by the breakage rate of ion pairs; the dependence of the “sticker” lifetime on salt concentration has been explored but is not well understood as yet. Direct ink writing additive manufacturing with a static mixer and fine-tip nozzle is studied by printing PDMS mixed either with fumed silica or as a two-part commercial liquid silicon rubber (LSR) mixed with polyethylene glycol (PEG). We assess their printability by printing a hollow slump cone, whose print quality is correlated with rheological measurements, including 1) a shear rate up-ramp followed by a down-ramp in shear rate, 2) creep tests at a series of increasing stresses, and 3) oscillatory shear with increasing amplitude well into the nonlinear regime. The PDMS-fumed silica mixtures fail to print even at the highest fumed silica loading used (9 wt.%), while LSR-PEG with 4 or 6 wt.% PEG prints very well even with low Shore hardness LSR. These large differences in printability of two classes of PDMS materials correlate poorly with rheological behavior in many of the above tests. The exceptions are the apparent yield stress during a down-ramp in shear rate following a previous up-ramp to the maximum shear rate of 1000 s-1, which is similar to the highest shear rate in the print nozzle, the stress at the crossover of the apparent G’ and G’’ curves in a strain amplitude sweep, and the stress at which irreversible flow becomes dominant in a stress sweep. These results indicate that the printability of the materials considered here depends strongly on both their yield stress and their ability to rebuild structure and yield stress quickly after experiencing the high shear rates characterizing their emergence from a narrow nozzle tip.

Advances in computer science technologies have shed light on artificial neural networks (ANN). ANN shows its power and efficiency in various classification and regression problems and is gaining more and more popularity in various fields. In this dissertation, we investigate treatment effects estimation using fully connected shallow network, and explore sparse deep neural network regression and functional derivative estimation. The first part of this dissertation provides a unified framework for efficient estimation of various types of treatment effects (TE) in observational data with a diverging number of covariates through a generalized optimization. We show that the number of confounders is allowed to increase with the sample size, and further investigate how fast it can grow with the sample size to ensure root-n consistency of the resulting TE estimator. Moreover, we establish asymptotic normality and semiparametric efficiency of the TE estimator. The second part of this dissertation proposes a penalized deep ReQU network estimator (PDRN) obtained from empirical risk minimization framework. The proposed neural network bases on Jacobi polynomial approximation on the hyperbolic cross/sparse grid and alleviates the "curse of dimensionality". Our PDRN estimator also provides smooth functional derivative estimation. Our estimators are illustrated through simulation studies and multiple real data examples.

This dissertation consists of three essays on historical and contemporary topics in health and human capital development. Chapter One examines the effects of the Great Depression on the educational outcomes of the 1918 influenza pandemic birth cohorts. I construct a novel dataset linking World War II draftees to the federal population censuses of 1920 - 1940 and newly digitized county-level data on pandemic severity. Using geographic variation in Great Depression exposure, pandemic severity, and their interaction, I find that the larger economic swings dampened the flu's negative effect on educational attainment. I present suggestive evidence that the underlying mechanism was the added-worker effect. Chapter Two collects and imputes data on the elementary and secondary school schedules, which can be linked to infectious disease patterns before and during the current SARS-COV-2/COVID-19 pandemic. Variation in school opening and closing dates during the academic year are shown to have been important drivers of infectious disease, consistent with schools constituting the primary locus of most social interactions among children. The literature on intergenerational transmission usually studies the top-down relationship from parents to their descendants. Chapter Three takes a backward approach and studies the effect of children’s education on parental longevity. Using the exogenous variation in the compulsory attendance laws, child labor laws, and continuation schooling laws in different states from 1880 to 1930 among different states, I estimate the causal effect of children’s education on parents’ longevity and mortality for the American Civil War veterans. To address the endogeneity of children's education, I instrument for actual years of schooling using required years of schooling. The instrumental variable is constructed from geographic variation in exposure to compulsory schooling, child labor, and continuation schooling laws.

This study aims to investigate the influence of the second language (L2) on the use of the first language (L1) in late bilinguals within an L1 dominant environment. Cross-linguistic influence (Kellerman & Smith, 1986) has been usually studied in the forward direction: how bilinguals’ L1 influences the acquisition and use of their L2. The other direction (i.e., the influence of L2 on L1), on the other hand, has not been sufficiently investigated. The current study looks at Chinese-speaking learners who acquire their L2 English through instruction in an L1 dominant environment. It does so by examining ‘subject realisation’, an area where Chinese and English exhibit substantial typological contrasts since Chinese allows both overt and null arguments under certain discourse-pragmatic conditions, whereas subjects in English are, under most circumstances, obligatorily expressed (Huang, 1984).. It is then hypothesized that long-time learning and regularly using English as L2 would increase the use of overt subjects realised in the bilingual’s first language, i.e., Chinese, with the consequent use of fewer null subjects in their L1. In addition, following Grosjean (1998), the interaction between the bilingual’s two languages is expected to be stronger when bilinguals produce language in the so called ‘bilingual mode’, i.e., when both languages are highly activated, than in a ‘monolingual mode’, i.e., when only one language is predominately activated. Such ‘language mode’ factor leads naturally to a futher hypothesis: fewer null subjects are realised in speech produced by Chinese-English bilinguals within a bilingual mode compared to monolingual mode.

Markov chain Monte Carlo (MCMC) methods are often used in Bayesian analysis to approximate expectation with respect to a certain distribution. Monte Carlo standard error (MCSE) can be used to determine the desired number of dependent samples, as well as to construct confidence intervals of MCMC estimates. It is usually unknown and various techniques have been suggested to estimate MCSE. A fundamental problem for these techniques is to choose an appropriate bandwidth. Previous research shows that a bandwidth proportional to $n^{1/3}$ is optimal for certain estimators. The proportional constant of $n^{1/3}$ however is unknown. As a result, $n^{1/3}$ is suggested although sub-optimal due to the missing proportional constant. $n^{1/2}$ was also considered to account for the constant but its asymptotic performance is worrisome.Existing literature mostly considered the above issues under univariate setting but Bayesian analysis normally involves multiple parameters. Computation time is a major challenge to estimate multivariate MCSE, where large amount of dependent samples are involved. Therefore multivariate estimators of MCSE that delivers fast and accurate estimation is desirable.This dissertation addresses the above two problems. I consider a family of estimators and established conditions under which their mean squared consistency exist. The results have a direct application in bandwidth selection and also suggests a bandwidth proportional to $n^{1/3}$. The proportional constant can be obtained based on the proof of mean squared consistency. I further suggest to approach the proportional constant with a pilot estimate. The suggested bandwidth shows superior performances compared with $n^{1/3}$ or $n^{1/2}$. The above results are established under multivariate setting which not only covers the long-standing univariate bandwidth selection problem, but also brings up the multivariate question with a solution. To tackle the computational problem in multivariate setting, I propose a family of new estimators and prove strong consistency of these estimators. The new estimators are fast to compute and have comparable performances to spectral variance estimators with slightly inflated variance.

Nanocomposites consist of graphene dispersed in matrices of soft materials are promising thermal management materials. A fundamental understanding of the thermal transport at graphene-soft material interfaces is essential for developing these nanocomposites. In this dissertation, thermal transport at graphene-octane interfaces was investigated using molecular dynamics simulations, and the results revealed several important characteristics of such thermal transport. The interfacial thermal conductance of graphene-octane interfaces were studied first. It was found that the interfacial thermal conductance exhibits a distinct duality: if heat enters graphene from one side of its basal plane and leaves it through the other side, the corresponding interfacial thermal conductance, Gacross, is large; if heat enters graphene from both sides of its basal plane and leaves it at a position far away on its basal plane, the corresponding interfacial thermal conductance, Gnon-across, is small. Gacross is ~30 times larger than Gnon-across for a single-layer graphene immersed in liquid octane. Additional analysis showed that this duality originates partially from the strong, positive correlations between the heat fluxes at the two surfaces of a graphene layer. The interfacial thermal conductance of the graphene-soft material interfaces in presence of defects in the graphene was then studied. The results showed that the heat transfer at the interfaces is enhanced by defects. Estimations based on effective medium theories showed that the effective thermal conductivity of the graphene-based composites could even be enhanced with defects in graphene when heat transfer at the graphene-soft material interface is the bottleneck for the thermal transport in these composites. To describe the interfacial thermal transport at graphene interfaces uniformly, a nonlocal constitutive model was proposed and validated to replace the classical Kapitza model. By characterizing the thermal transport properties of graphene interfaces using a pair of thermal conductance, the model affords a uniform description of the thermal transport at graphene interfaces for different thermal transport modes. Using this model, the data interpretation in time domain thermalreflectance (TDTR) measurements was investigated, and the results showed that the interfacial thermal conductance measured in typical TDTR tests is that of the across mode for thin-layered materials.

This thesis focuses on determination and correction of the information about disturbances obtained from images taken through a turbulent medium. The more distorted the images of a scene, the more corrections related to disturbances can be utilised to obtain unperturbed images. Two image restoration approaches explored in this thesis are: the Bispectrum algorithm and Lucky Region method. Also, an Image Alignment method is used as a pre-processing approach for correcting random shifts in an image caused by refractive-index variations in a turbulent atmosphere. The improved Bispectrum algorithm is applied in association with a speckle imaging method. Speckle imaging methods process a large number of short-time-exposure images. The turbulence in each of the short-time-exposure images is considered a frozen pattern. In the work of Wen [2], on which this work is based, there was an outstanding question of what sub- region size to choose to gain the preferable result. This work investigates this further in an empirical fashion, and examines whether the choice is image independent. A further examination of its properties shows that, when a 16×16 sub-region size is used with taking around 80 image frames, the best image restoration result is obtained. The Lucky Region method attempts to select the higher-quality regions from all image frames based on the value of Bicoherence, an image quality indicator, introduced by Wen [2] and shown to be effective for Lucky Region selection. An investigation of the range of ratios needed for the threshold in this method is provided. The best restored image occurs for threshold values defined between 10% and 20% of the frames with a 64×64 sub-region size. To achieve a better image reconstruction result, Image Alignment, which seeks to reduce the effect of the scene movement in an image sequence to prepare a better database for further image operations, is introduced for application in association with the Bispectrum and Lucky Region method. These combined image restoration algorithms demonstrate better performances than the direct approach of Wen’s original algorithms and are considerably more computationally efficient.

The challenge for contemporary literary scholars comes down to the question of how useful a cultural product is in describing and configuring the contemporary cultural scene, for example, the impact of literature, film, and TV in the globalizing era. Many scholars, in order to capture this diversified post-socialist cultural phenomenon, explore as many genres of art as they can in their research. However, identifying the common feature of post-socialist diversity in China still remains a question without answer. My research focuses on the theme of youth narratives in order to provide a solid and concrete perspective to assess contemporary Chinese culture and society.My research is motivated by a cultural phenomenon that I observe in contemporary China, namely the prevalence of youth narratives in fiction, film, and popular culture. The youth narrative has been frequently intertwined with China’s pursuit of modernity and its nation-building projects under different political and cultural ideologies. My research traces the emergence and evolution of youth narratives in modern Chinese literature and then delves into various recurring themes and motifs about youth in literature, film, and TV from the 1990s to 2010, to discuss the cultural, social and political significances of youth narratives in contemporary China in order to capture common and critical features of diverse post-socialist status and engage with the furious debate on post-socialism and globalization in the Chinese context. What is more, my study reveals conflict between elitist state discourse and civil discourse implied by youth narratives from the 1990s to the present. This period provides rich textual and visual materials that showcases the most heatedly-discussed phenomena in China and resonates with classic dilemmas and ambivalences of Chinese modernity. Moreover, each of these phenomena relates to drastic economic change and social transition during the contemporary period. In a nutshell, these texts do not only reflect the moment but are also part of the moment. Hence, my research of contemporary Chinese culture links the past, present, and future; and is located at the intersection of historical and cultural studies, providing both diachronic and synchronic visions of key issues such as modernity, nationalism, and globalization.

Instruction in Chinese as a foreign language has historically been focused on the beginning and intermediate levels of proficiency; however, as a field we have just now begun to address the advanced levels of instruction where the ability to function in professional environments in China is a major goal. With the rise of the Internet and the fact that many young Americans are employed in Chinese work environments, it is time for serious research into how composition instruction in Chinese as a foreign language can help students create identifiable cultural performances. In order to highlight the concept of readership in the CFL field of Chinese L2 composition, this dissertation examines how a group of American undergraduate students interpreted evidence as they took part in an intensive study abroad program to learn Chinese persuasive composition. Participants in the study included nine Chinese L2 students from five American universities and the course instructor. I utilized various data collection method, such as questionnaires, classroom observation, field notes, audio recordings, document gathering, and (text-based) interviews. Results of the study suggest that while both the students and the textbook/instructor viewed evidence as the materials supporting the writer’s opinions, the actual student interpretations of evidence varied greatly from those presented in class. We found that what is convincing varies according to cultural background. The purpose of persuasive composition in L2 classes should, be to reach the reader in the L2 culture, rather than to reaffirm the writer’s positions in his or her base culture. This study highlights the need for Chinese L2 composition teachers to reexamine the role of the reader, which has been neglected in Chinese L2 composition classes. Readership is the result of certain socio-cultural interactions. As a greater number of Americans become part of the Chinese workforce, designers of L2 curricula have the formidable task of identifying those composition tasks that will be most beneficial to a foreigner’s career in China and developing efficient pedagogical approaches that prepare CFL students to meet their readers’ expectations while accomplishing those tasks. Chapter 4 analyzes the results of a survey of 51 Chinese L2 learners who were working in China or doing academic research there in the summer of 2012 or before. In this study we identified the composition performances typically found in professional and academic contexts in China and analyzed how the concepts of genre and genre colonies might be utilized to describe L2 writer’s composition performances in L2 communities for pedagogical purposes. In Chapter 5, we propose a performed-culture approach in Chinese L2 composition instruction. My intention is that this new approach will help Chinese L2 students realize professional goals by producing texts that meet the expectations of their Chinese readers.

Falls are a prevalent and significant problem facing a growing older population. Many existing clinical fall risk assessment tools are complicated, time-consuming, and involve subjective input from either clinical practitioners or the older individuals being assessed. In addition, most clinical assessment tools have undergone only limited independent validation to ascertain their true accuracy in predicting future falls. Body-worn sensor-based movement analysis shows great promise as a means of assessing fall risk, given that it enables quantitative signal analysis and can be integrated into a simple assessment procedure. Such body-worn sensor-based systems could also feasibly be used for unsupervised assessment and long-term monitoring. Previous studies within the author's research group have attempted to construct a waist-worn accelerometry-based system, using time domain features extracted from the accelerometry signals, which were acquired from the waist-worn triaxial accelerometer (TA) as the study participant executed a scripted movement routine. These studies aimed to develop a model which could estimate fall risk for the 68 elderly subjects, by mapping to a clinical fall risk assessment score. Based on this work, this thesis report firstly presents an attempt to improve this model, using spectral analysis of the same accelerometry signals. However, a strong optimistic bias was discovered in how validation (estimating future model performance) was previously performed for this modelling technique; the modelling method is adjusted to remove this validation bias. Another major challenge encountered relates to the large feature set dimensionality and small available sample size, hence a dimensionality reduction method is employed prior to model training. Subsequently, as one of its major contributions, this thesis provides an external validation of this waist-worn sensor-based fall risk assessment system using an independent group of subjects recruited into a validation study. However, the results of this validation show poor performance in estimating the reference clinical fall risk test scores which the proposed sensor-based fall risk assessment model attempts to estimate. This thesis project also presents a Poisson regression model to estimate the fall rates of the subjects over a period of several months (up to 13 months) following an initial sensor-based assessment, using the same group of accelerometry-based features from the proposed directed movement routine. An independent validation is also performed using assessment data from the same validation study; however, the validation performance is also poor. A discussion of the necessity and disadvantages of the dimensionality reduction process used prior to modelling is raised, highlighting how some useful features are discarded during this procedure, but that this process is entirely necessary in order to mitigate the risk of model over-fitting. Subsequently, a TA-based reaction time (RT) test and a new walking task (the six meter walk test (6MWT)) are investigated during the validation study, and exploratory evaluations of the TA-based RT test and the 6MWT in model training for this study cohort are presented. The TA-based aural stimulus-based RT test is moderately correlated with a better-known clinical vision-based RT test, and performs worse than the clinical RT test in prospective fall prediction model training. Compared to other movement tasks investigated, the use of 6MWT features alone provided the best training performance in prospective fall prediction, with RMSE=0.80 and R2=0.66. This thesis puts forward the view that sensor-based gait and movement assessment for fall risk assessment continues to show promise, however, two major challenges face this field of research, which are currently not widely acknowledged. The first relates to sample size and the curse of dimensionality; given the rich information contained within inertial signals captured during assessment, the number of assessments available (particularly those with prospective fall count follow-up data) is usually exceptionally small by comparison, which leads to a great risk of detecting false positive relationships between some of the many extracted inertial features and the target (future falls). The second major issue relates to validation. Very few sensor-based assessment techniques perform an independent external validation, but rather rely on internal validation (using the cross validation method), or simply present the results of training. External validation is a critical step in statistical model development and should always be performed before a model's value can be asserted with any confidence. To overcome these challenges to ongoing fall prediction research efforts, greater global coordination is needed. In order to accumulate large enough sample sizes to avoid making type I errors when selecting predictive features, to complete the task of gathering prospective fall count records, and to execute external validation studies of proposed models, a concerted and coordinated global effort is required. Given such a collaborative effort, there is hope that a better understanding of the causes of falls may be reached, allowing specific interventions to be targeted at these deficits, and ultimately remedy the cost and suffering associated with falling.

Surface plasmon resonance (SPR) has been widely used as a powerful analytical technique for the study of a broad range of biomolecular interactions. With the capability of real-time detection, SPR allows convenient and nondestructive measurement of analyte concentration and binding kinetics. To improve the performance of SPR biosensing, we have developed a series of novel methods that lead to ultrasensitive detection via signal amplification by coupling inline atom transfer radical polymerization (ATRP) to a biospecific interaction. To adapt this enhancement method to the fragile lipid membrane-based system, initiator functionalized gold nanoparticles (AuNPs) has been used with in situ ATRP reaction, which allows sensitive measurement of lipid membrane binding proteins. In addition, conjugated nanoparticles AuNPs and Fe3O4 NPs have been employed to enhance protein detection sensitivity with SPRimaging. A novel Au-well microarray, fabricated based on spatial variation of metal thickness, is utilized to eliminate background resonance and provide better performance. To mimic cell membrane recognition processes, a membrane bilayer has been assembled onto a nanoglassified surface, and a synthesized water-soluble deep cavitand was incorporated into the supported lipid membrane. The incorporated cavitand retains its host properties and real-time analysis of the host:guest properties is carried out by SPR and fluorescence microscopy. To further explore the biomolecule interactions on cell membrane, we have studied the polymer growth at the lipid membrane-water interface using functional guest initiator that triggers an ATRP process. A variety of functional polymers have been grown at the lipid membrane surface, demonstrating different hydrophilicity and protein binding properties. Mass spectrometry in combination with SPR for comprehensive protein analysis has been explored to build a new technical platform. We have designed a thermoresponsive poly(N-isopropylacrylamide) (PNIPAAM) grafted surface, which demonstrates temperature-mediated hydrophobicity for enrichment of proteins based on hydrophobic interaction. SPR quantitatively measures the binding of proteins and peptides, and the biochip is directly interfaced with matrix assisted laser desorption/ionization (MALDI)-MS to characterize the bound analytes. It provides outstanding results in simple steps of on-plate desalting, offering news tools for effective proteomic analysis.

It was generally recognized that the bioavailability, other than total content of heavy metals in soil, is crucial to predict metals' toxic effect on environment. Stable Isotope Dilution technique (SID) equipped with inductively coupled plasma-mass spectrometry (ICP-MS) is a good way for measuring the potential bioavailable portion of metals in soils. In Chapter 2, we build method to accurately determined Cd (111Cd/114Cd) and the ratio of Zn (66Zn/68Zn) in soil suspension and then applied this to stable isotope dilution technique to successfully obtain the labile fraction of Cd and Zn in polluted soils. In Chapter 3, it was found when compared to chemical extraction methods, the stable isotope dilution technique is the better way to assess bioavailability of Cd and Zn in soil to the earthworms Eisenia fetida and Lumbricus terrestris. Besides, the result showed on average, the fraction of labile Cd (44.7%) was significantly higher than that of Zn (25.4%), indicating that Cd is more chemically reactive in soil, which may explain the larger BSAFs for Cd in earthworms. In addition, it was found the accumulation of heavy metals was not only element-specific, but also species-specific. In Chapter 4, the uptake and elimination kinetics of Zn and Cd by E. fetida and L. terrestris were investigated then. The kinetic rates were calculated using a first-order, one-compartment, toxico-kinetic model. Pronounced differences were found on the earthworms' uptake and elimination kinetics between Zn and Cd (p

In recent decades, the wireless communication industry has attracted a great deal of research efforts to satisfy rigorous performance requirements and preserve high spectral efficiency. Along with this trend, I/Q modulation is frequently applied in modern wireless communications to develop high performance and high data rate systems. This has necessitated the need for applying efficient complex-valued signal processing techniques to highly-integrated, multi-standard receiver devices. In this dissertation, novel techniques for complex-valued digital signal enhancement are presented and analyzed for various applications in wireless communications. The first technique is a unified block processing approach to generate the complex-valued conjugate gradient Least Mean Square (LMS) techniques with optimal adaptations. The proposed algorithms exploit the concept of the complex conjugate gradients to find the orthogonal directions for updating the adaptive filter coefficients at each iteration. Along each orthogonal direction, the presented algorithms employ the complex Taylor series expansion to calculate time-varying convergence factors tailored for the adaptive filter coefficients. The performance of the developed technique is tested in the applications of channel estimation, channel equalization, and adaptive array beamforming. Comparing with the state of the art methods, the proposed techniques demonstrate improved performance and exhibit desirable characteristics for practical use. The second complex-valued signal processing technique is a novel Optimal Block Adaptive algorithm based on Circularity, OBA-C. The proposed OBA-C method compensates for a complex imbalanced signal by restoring its circularity. In addition, by utilizing the complex iv Taylor series expansion, the OBA-C method optimally updates the adaptive filter coefficients at each iteration. This algorithm can be applied to mitigate the frequency-dependent I/Q mismatch effects in analog front-end. Simulation results indicate that comparing with the existing methods, OBA-C exhibits superior convergence speed while maintaining excellent accuracy. The third technique is regarding interference rejection in communication systems. The research on both LMS and Independent Component Analysis (ICA) based techniques continues to receive significant attention in the area of interference cancellation. The performance of the LMS and ICA based approaches is studied for signals with different probabilistic distributions. Our research indicates that the ICA-based approach works better for super-Gaussian signals, while the LMS-based method is preferable for sub-Gaussian signals. Therefore, an appropriate choice of interference suppression algorithms can be made to satisfy the ever-increasing demand for better performance in modern receiver design.

Biochemical activity and core stability are essential properties of proteins, maintained usually by conserved amino acids. Structural dynamics emerged in recent years as another essential aspect of protein functionality, which enables the adaptation of the protein to substrate binding. It also underlies its ability to undergo allosteric transitions, while maintaining its fold. Key residues that mediate structural dynamics would thus be expected to be conserved, or exhibit co-evolutionary patterns at least. Yet, the correlation between sequence evolution and structural dynamics is yet to be established. To this end, we have performed in-depth analyses of a number of representative proteins, using a combined approach of sequence analyses and coarse-grained physics-based models. For the Hsp70 family, we studied the interactions of Hsp70 ATPase domains with four different nucleotide exchange factors (NEFs) and revealed two classes of key residues: (i) those highly conserved residues involved in nucleotide binding, which mediate the ATPase domain opening via a global hinge-bending, and (ii) those co-evolving and highly mobile residues engaged in specific interactions with NEFs. The observed interplay between these respective intrinsic (pre-existing, structure-encoded) and specific (co-evolved, sequence-dependent) interactions provides us with insights into the allosteric dynamics and functional evolution of the modular Hsp70 ATPase domain, and inspired a follow-up study that identified a group of key residues mediating the Hsp70 allosteric pathways using perturbation analysis. Along the same lines, a systematic study has been performed on a set of 34 enzymes representing various folds and functional classes, which generalizes the previous findings and unravels a unique correlation between sequence evolutionary properties and conformational dynamics. Our findings suggest that there is a balance between physical adaptability (enabled by structure-encoded motions) and chemical specificity (conferred by correlated amino acid substitutions), and this balance underlies the selection of a relatively small set of versatile folds by proteins. In another study, HIV-1 protease was investigated as a special case in which short-term evolutionary pressure plays a significant role. With advanced clustering techniques, we differentiated multi-drug resistant mutations from those arising from phylogenetic variations; correspondingly, these mutations exhibit distinctive structural/dynamical features, underlying the role of protein dynamics in conferring drug resistance.