Master of Science (MS), Ohio University, 2012, Civil Engineering (Engineering and Technology)

Many variable stiffness and damping devices have been proposed in an attempt to mitigate the effects of unwanted vibrations in civil engineering structures. This work proposes a novel variable stiffness or damping device called the Continuously Variable Amplification Device (CVAD) that improves upon past limitations. It consists of a sphere and rollers in series with a spring or damper, and is able to produce a large, continuous, and rapidly varying range of stiffness or damping values. A series of relationships and equations are derived and validated to describe the amplification of the device. A numerical simulation is developed that pairs the CVAD with SSASD and RSASD devices and places it in a seismically excited eight story structure in a supplemental damping role. In addition, centralized control laws are created and applied to the CVAD devices. The results demonstrate the device and its centralized control laws are a marked improvement over the state-of-the-art.

Doctor of Philosophy, The Ohio State University, 2023, Aero/Astro Engineering

The feasibility of an active flow control enabled variable area turbine was explored. Pressurized air was ejected from the nozzle guide vanes to reduce the effective choke area, and mass flow rate through, the turbine inlet. A set of experimental and computational studies were conducted with varying actuator types and parameters to determine their effectiveness and develop models of the flow physics. Preliminary results from a simple quasi-1D converging-diverging nozzle, with an injection flow slot upstream of the throat, showed a 2.2:1 ratio between throttled mass flow rate and injected mass flow rate at a constant nozzle pressure ratio. The penetration of the injection flow and corresponding reduction in the primary flow streamtube were successfully visualized using a shadowgraph technique. Building on this success, a representative single passage nozzle guide vane transonic flowpath was constructed to demonstrate feasibility beyond the quasi-1D converging-diverging nozzle. Both secondary slot blowing from the vane pressure surface and vane suction surface just upstream of the passage throat again successfully reduced primary flow. In addition, fluidic vortex generators were used on the adjacent suction surface to reduce total pressure loss along the midspan and further throttle the primary flow. Computational fluid dynamics simulations were used to explore the effects of a variety of parameters on the flow blockage and actuator effectiveness. Simplified models were developed to describe the relationships of various factors impacting flow blockage, turning angle, and total pressure loss. Finally, the active flow control systems were simulated at engine relevant pressures and temperatures and found to have only a minimal drop in total pressure recovery and effectiveness, which could be predicted by the simplified blockage model.

Master of Science (MS), Bowling Green State University, 2022, Physics

Globular cluster M5 (NGC 5904), located in the constellation Serpens, was discovered over 300 years ago. Since then, several studies have sought to characterize and understand its origin and evolution. Of particular interest are the studies involving the cluster's variable star population. To date nearly two hundred variable stars have been identifed in the cluster. The majority of these are short period variable stars belonging to the category “RR Lyrae”. With advances in photometry and automated observation, recent studies have begun to investigate the cluster's population of long period variable stars (LPV's), but further study is needed to fully catalog and characterize them. Seven of the twenty known LPV's in M5 do have undetermined periods, and more are awaiting discovery. With the use of data acquired with the BGSU half-meter telescope and from the Panchromatic Robotic Optical Monitoring and Polarimetry Telescopes (PROMPT), period analysis was performed on M5's two II Cepheids, 20 known LPV's and one newly discovered LPV. The creation of a color-magnitude diagram (CMD) for M5 allows for testing the hypothesis that as a star evolves further along the red giant branch of the CMD, that the star's periodicity becomes more regular, and will show a larger amplitude of variability. The positions of the stars of interest, and the conclusions drawn from their light curves in our data was compiled and the variability type for each star was assessed. Suggestions for future observational work are provided with the aim of further improving upon the characterization of these LPV's

Master of Science (MS), Bowling Green State University, 2018, Physics

Photographic plates in the Harvard collection show the star V1835 Aql brightening by a factor of 100 in flux over four years starting in 1899, remaining at maximum for four years, then declining below the depth of the plates \cite{williams2005}. This nova-like behavior is very atypical for most variable stars and as a result there has been much debate over the exact nature of V1835 Aql. This debate was ended by the discovery of a Raman scattered emission line at 6824 \AA, which is unique to symbiotic binaries and unequivocally identifies V1835 Aql as a symbiotic star \cite{bk12}. Our research hopes to expand upon our knowledge of V1835 Aql through analysis of five years worth of multi-band optical time-series photometry. From this we have found the period of this star to be 419 days. This long period confirms that V1835 Aql is a symbiotic star and not its closer orbiting cousin, a cataclysmic variable. We have also determined the properties of all the other variable star candidates near V1835 Aql, of which there are 31.

Master of Science (MS), Ohio University, 2017, Mechanical Engineering (Engineering and Technology)

Moving water quickly and efficiently has always been crucial to human development, especially in agriculture. However, this challenge still often goes unmet, especially in the developing world where access to infrastructure and mechanized pumping equipment is limited. In such locations human power is the most readily available source of power. The Beale Continuously Variable Transmission (CVT), modeled and experimentally examined by Cyders (2012), has potential to address this challenge. This thesis will adapt the work done by Cyders to a slider-crank simulation paired with a positive displacement pump. Predictions of system flowrate response at varying combinations of pressure and input shaft speed will be made using this model, and will then be compared against experimental results taken from a physical prototype of the system. Comparisons between these data sets will form the final result of this project, and will inform recommendations for any future work.

Doctor of Philosophy, The Ohio State University, 2016, Biostatistics

In functional genomics studies, human tissue samples are always difficult to get access to, and the lab experiments are expensive to implement and time-consuming. Data mining in existing databases is an essential step in building scientific hypotheses for designing well-targeted lab experiments. Therefore, it is important to study statistical methods that can better utilize observational data in functional genomics studies. Measuring allele-specific RNA expression provides valuable insights into cis-acting genetic and epigenetic regulation of gene expression. Widespread adoption of high throughput sequencing technologies for studying RNA expression permits measurement of allelic RNA expression imbalance at heterozygous single nucleotide polymorphisms (SNPs) across the entire transcriptome, and this approach has become especially popular with the emergence of large databases, such as GTEx. However, the existing methods used to model allelic expression from RNA-seq often assume a strong negative correlation between reference and variant allele reads, which may not be reasonable biologically. In Chapter 2, a folded Skellam mixture model is proposed for AEI analysis using RNA-seq data. Under the null hypothesis of no AEI, a group of SNPs (possibly across multiple genes) is considered comparable if their respective total sums of the allelic reads are of similar magnitude. Within each group of comparable SNPs, we identify SNPs with AEI signal by fitting a mixture of folded Skellam distributions to the absolute values of read differences. By applying this methodology to RNA-Seq data from human autopsy brain tissues, we identified numerous instances of moderate to strong imbalanced allelic RNA expression at heterozygous SNPs. Findings with SLC1A3 mRNA exhibiting known expression differences are discussed as examples. In the theory of complex systems, the Sobol sensitivity indices are typically introduced under the high dimension model representation (HDMR, also known (open full item for complete abstract)

Master of Science in Electrical Engineering, Cleveland State University, 2014, Washkewicz College of Engineering

Automotive system optimization problems are difficult to solve with traditional optimization techniques because the optimization problems are complex, and the simulations are computationally expensive. These two characteristics motivate the use of evolutionary algorithms and meta-modeling techniques respectively. In this work, we apply biogeography-based optimization (BBO) to radial basis function (RBF)-based lookup table controls of a variable camshaft timing system for fuel economy optimization. Also, we reduce computational search effort by finding an effective parameterization of the problem, optimizing the parameters of the BBO algorithm for the problem, and estimating the cost of a portion of the candidate solutions in BBO with design and analysis of computer experiments (DACE). We find that we can improve fuel economy by 1.7% compared to the original control parameters, and we find effective, problem-specific values for BBO population size and mutation rate. Finally, we find that we can use a small number of samples to construct DACE models, and we can use these models to estimate a significant portion of the BBO candidate solutions each generation to reduce computation effort and still obtain good BBO solutions.

Doctor of Philosophy, The Ohio State University, 2008, Statistics

We consider statistical procedures for feature selection defined by a family of regularizationproblems with convex piecewise linear loss functions and penalties of l1 or l∞ nature. For example, quantile regression and support vector machines with l1 norm penalty fall into the category. Computationally, the regularization problems are linear programming (LP) problems indexed by a single parameter, which are known as “parametric cost LP” or “parametric right-hand-side LP” in the optimization theory. Their solution paths can be generated with certain simplex algorithms. This work exploits the connection between the family of regularization methods and the parametric LP theory and lays out a general simplex algorithm and its variant for generating regularized solution paths for the feature selection problems. The significance of such algorithms is that they allow a complete exploration of the model space along the paths and provide a broad view of persistent features in the data. The implications of the general path-finding algorithms are outlined for various statistical procedures, and they are illustrated with numerical examples.

Doctor of Philosophy, The Ohio State University, 2003, Industrial and Systems Engineering

This dissertation re-examines some of the most basic design of experiment methods with respect to their ability to achieve intuitive objectives. For example, it provides probably the first comprehensive evaluation of the ability of standard screening approaches to correctly tell which factors have important effects on average outputs. Also, the dissertation examines the prediction errors that users of so-called mixture experimental design and qualitative response surface methods can achieve. In practical situations, the derived "decision support" information can tell the user in advance whether the number of runs used is adequate for the experimenter's needs and provide a basis for selecting one method over another when alternatives are presented. Also, the dissertation clarifies, perhaps for the first time, the potentially serious prediction error issues associated with the methods that have been proposed for response surface investigation when some factors are qualitative. In addition to developing comprehensive computational studies of existing methods, new methods are proposed with potentially important advantages. For example, the dissertation provides some of the first unbalanced screening experimental plans relevant to cases in which some combinations of settings have far higher costs than other combinations. For situations in which some factors are mixture components, e.g., %CO2, %Ar, %N, and other factors are process variables, the dissertation provides some of the first economically relevant experimental plans offering potentially substantial reductions in prediction errors. Also, the dissertation provides the first truly advisable experimental designs for many response surface cases in which some variables are qualitative. All new methods are derived from optimization formulations or "improvement systems design problems". In each case, the intent is to design the method using the objective or objectives that most directly describe the purpose of the impro (open full item for complete abstract)

Master of Science (MS), Bowling Green State University, 2011, Physics

Studying the late stages of stars is essential to understand the process of stellar evolution. Depending on their masses and properties, some stars become unstable at the end of their evolutionary state and hence they start pulsating. Their brightness and surface temperature change with their pulsations and hence we see them as variable stars. We are looking for long period variable stars (LPVs) in the globular cluster NGC 6496. We observed the cluster from February 2009 till October 2010 using a 0.41m telescope in the V and I bandpasses. We have identified 11 variable stars in the cluster. 6 of them are new discovered LPVs in which 3 of them are semiregular LPVs and the rest 3 are irregular LPVs. We plotted the color magnitude diagram (CMD) of this cluster and all our LPVs were detected on the RGB/AGB. 5 of the 11 variable stars are short period variable stars in which 4 of them are W UMa binary stars and 1 is an Algol binary star. The light curves of all these stars are plotted in this paper and the periods were detected using different period-finding methods.

Master of Science (MS), Bowling Green State University, 2011, Physics

We have searched for Long Period Variable (LPV) stars in the metal-rich cluster NGC 6388 using time series photometry in the V and I bandpasses. A CMD was created, which displays the tilted red HB at V = 17.5 mag. and the unusual prominent blue HB at V = 17 to 18 mag. Time-series photometry and periods have been presented for 63 variable stars, of which 30 are newly discovered variables. Of the known variables nine are LPVs. We are the first to present light curves for these stars and to classify their variability types. We find 3 LPVs as Mira, 6 as Semi-regulars (SR) and 1 as Irregular (Irr.), 18 are RR Lyrae, of which we present complementary time series and period for 14 of these stars, and 7 are Population II Cepheids, of which we present complementary time series and period for 4 of them. The newly discovered variables are all suspected LPV stars and we classified them, using time series photometry and periods, as Mira for 1 star, SR for 15 stars, Irr for 7 stars, Suspected Variables for 7 stars, out of which there are 3 very bright stars that could have overexposed the CCD, with no definite borderline between the SR and Irr stars. Once classified we used probable distance for the cluster center and location on the CMD to establish possible membership, which left us with 63 possible cluster members, but the crowdedness of the cluster and the fact that the cluster is located near the bulge of the Milky Way prevents us from establishing a better certainty for its membership.

Master of Science (MS), Bowling Green State University, 2010, Physics

Photometry for the metal-rich globular cluster NGC 6553 has been done and is presented in both V and I banpasses. A color magnitude diagram for this cluster was created which indicates the presence of two distinct giant branches that separate the cluster population from the field population at I = 12.2 mag. Time-series photometry is presented for 49 stars, 22 of which are considered cluster members. Cluster membership was found by locating them on the cluster giant branch. Ten of these 22 cluster members are new discoveries while the other twelve were found during a study done by Lloyed Evans and Menzies in 1972. Where possible, properties, such as average period, average amplitude, and the observed range of magnitudes of the variable star candidates are discussed. Lloyd Evans and Menzies did not investigate the stars' properties, but they provided generalized sub-classifications for them. Their findings included eleven irregular/semi-regular variables and one Mira variable. We, on the other hand, provide specific sub-classifiactions for the individual stars. Comparing our results of the same twelve stars to Lloyd Evans and Menzies', we find that we sub-classified them as ten irregular/semi-regular variables, one Mira variable, and one possible binary star system. We do not find a definite borderline between the irregular and the semi-regular variable star classes. Also, we have made the observation that there is no trend of sub-classes with respect to the position on the giant branch. This leads to the conclusion that there is no smooth transition from the position of irregular/semi-regular variables to the position occupied by Mira-type variables, but rather a sudden jump. This same phenomenon has been observed by Lebzelter and Wood in a paper published in 2005.

Doctor of Philosophy (PhD), Ohio University, 2012, Mechanical Engineering (Engineering and Technology)

Efficient, high-performance continuously variable transmission (CVT) technology is currently being pursued by many engineering companies as a way to combat manufacturing cost, improve the feasibility of systems such as electric cars and small-scale wind power generation, and improve efficiency of many mechanical systems such as industrial pumps and fans. Successful, wide-scale implementation of efficient CVT technology has the capacity to reduce global energy consumption, diversify feasible energy sources and improve production and operating cost of many mechanical systems, but most contemporary designs have worse efficiency, cost or performance than the systems they are meant to replace. A new, unique mechanism called the Beale CVT has the capacity to overcome all these obstacles, but presents difficulties in design based on its dynamic characteristics, which have not previously been modeled or verified through experiment. This work successfully developed an accurate dynamic model for the elements making up the Beale CVT to be used in future design of the mechanism, which was then verified by experiment on a physical transmission prototype.

Doctor of Philosophy, Case Western Reserve University, 2025, Epidemiology and Biostatistics

Variable selection in high-dimensional data analysis poses substantial methodological challenges. While numerous penalized variable selection methods and machine learning approaches exist, many demonstrate instability in real-world applications. This thesis makes two primary contributions: developing a novel ensemble algorithm for variable selection in competing risks modeling and conducting a comprehensive stability analysis of established variable selection methods. The first component introduces the Random Approximate Elastic Net (RAEN), an innovative methodology that offers a stable and generalizable solution for large-p-small-n variable selection in competing risks data. RAEN's flexible framework enables its application across various time-to-event regression models, including competing risks quantile regression and accelerated failure time models. We demonstrate that our computationally-intensive algorithm substantially improves both variable selection accuracy and parameter estimation in a numerical study. We have implemented RAEN in a user-friendly R package, freely available for public use. To demonstrate its practical utility, we apply RAEN to a cancer study, successfully identifying influential genes associated with mortality and disease progression in bladder cancer patients. The second component comprises a systematic evaluation of eight variable selection methods' stability under varying conditions. Through comprehensive numerical studies, we examine how factors such as sample sizes, number of predictors, correlation levels, and signal strength influence performance. Based on these findings, we provide evidence-based recommendations for implementing variable selection methods in real-world data analysis.

PhD, University of Cincinnati, 2024, Arts and Sciences: Statistics

This dissertation explores Bayesian methods in modeling count data with excessive zeros through generalized linear mixed-effect models. It focuses on developing efficient computational algorithms, modeling shared independent and dependent random effects, and implementing effective variable selection. In Bayesian modeling of the zero-inflated count data, a challenge lies in the lack of closed-form posteriors and thus efficiency in Markov chain Monte Carlo (MCMC) samplings. We implement a data augmentation strategy based on Polya-Gamma latent variables. Under this approach, the binomial likelihood is represented as a Gaussian mixture with regard to the Polya-Gamma latent variables, leading to closed-form posteriors for regression coefficients. This greatly facilitates the computation related to the binomial likelihood, which is needed in modeling the zero-inflation using the logistic regression model as well as the over-dispersion using the negative binomial regression. In Chapter 2, we propose several Zero-Inflated Poisson (ZIP) and Zero-Inflated Negative Binomial (ZINB) models to address the unique challenges which are commonly observed in ecological studies, including overdispersion, excess zeros, missing data, and temporal dependencies. These models include shared year effects and autoregressive random components to capture both within- and between-year correlations. Simulation studies and real-world applications show that including autoregressive components in zero-inflated count models significantly improves model performance in modeling temporally dependent count data, particularly in terms of predictive accuracy and parameter estimation. The application of these models is shown by exploring migration patterns and potential environmental factors of a steelhead smolt migration dataset from Southern California Santa Clara River. The results show that models with autoregressive components have better performance than the other models in both model fi (open full item for complete abstract)

Doctor of Philosophy (Ph.D.), Bowling Green State University, 2024, Statistics

The sequential linear model is widely employed to analyze the dynamic data where the response variable at each time point incorporates the lagged results from the previous time point. With the lagged dependent response variables added to the model longitudinally, the issue of multicollinearity arises. In such situations, the Lasso method proposed by Tibshirani (1996) addresses both parameter estimation and variable selection simultaneously. However, in high-dimensional data and multicollinearity, the Lasso method can introduce bias in coefficient estimation and inconsistency in variable selection. To improve the Lasso method, a number of different penalty terms are proposed. Among the Lasso methods with different penalty terms, selecting an appropriate estimation and variable selection method is challenging work because it requires balancing the trade-off between achieving low bias and maintaining high prediction accuracy. One of the primary inferences in the sequential linear model is to predict the response variable with high accuracy and relatively minimal prediction errors, thereby saving time and expenses. To achieve this goal, we propose the estimation and variable selection method based on the Lasso, named Smoothly Clipped Absolute Deviation Penalty (SCAD) (Fan and Li, 2001), in the sequential linear model. The proposed SCAD method performs effectively in parameter estimation with low bias and variable selection with low predicted errors. In the demonstration of the effectiveness of the proposed method, we conduct the simulations where we compare the SCAD method with other methods including the ordinary least squares (OLS), Lasso, and Adaptive Lasso in both linear regression and sequential linear models. Since time series refers to a sequence of data generated at each time point, where the lagged response variable at each time point is used as a predictor in the subsequent time point model, accounting for errors based on assumptions, we simulate the data in (open full item for complete abstract)

Master of Science, University of Toledo, 2024, Chemistry

Thermal expansion is a physical property that may contribute to materials' malfunctioning in applications ranging from various electronics to construction and other engineering fields. As heat is applied to or inherently generated by materials, they tend to expand, thereby causing stress, strain, cracks, and structural distortion at the interfaces between dissimilar materials. These structural misalignments, resulting from thermal expansion, adversely affect the properties of a material, which in turn leads to a change in a material's performance. This change in performance may disrupt the original purpose for which the material was made. These challenges make complementary materials that can reduce or eliminate the thermal expansion of other materials when incorporated into a composite attractive. Negative thermal expansion (NTE) materials are materials that contract upon heating. These materials can serve as fillers in composites to complement positive expansion materials and reduce overall thermal expansion in composite materials. Such composites can find applications in high precision optical mirrors, in the aerospace industry, in dental fillings, and ultimately, in various electronics. However, a thorough investigation of these promising materials is needed to understand some of the problems currently preventing full implementation. Among these challenges, avoiding temperature and pressure induced phase transitions that form positive expansion polymorphs has been an important factor. These phase transitions destroy the NTE property of the materials. Hence, stabilizing the NTE phase in a wider temperature and pressure range will enhance the materials' potential applications. This research focuses on the scandium tungstate (Sc2W3O12) family of NTE materials, represented as A2M3O12 (A = trivalent cation, M = tungsten, molybdenum). This family was chosen because of the wide range of cations that can be incorporated into the structure due to the chemical flexibil (open full item for complete abstract)

Master of Science in Electrical Engineering (MSEE), Wright State University, 2024, Electrical Engineering

This research illustrates a high-security wireless communication method using a joint radar/communication waveform, addressing the vulnerability of traditional low probability of detection (LPD) waveforms to hostile receiver detection via cyclostationary processing (CSP). To mitigate this risk, RF steganography is used, concealing communication signals within linear frequency modulation (LFM) radar signals. The method integrates reduced phase-shift keying (RPSK) modulation and variable symbol duration, ensuring secure transmission while evading detection. Implementation is validated through software-defined radios (SDRs), demonstrating effectiveness in covert communication scenarios. Results include analysis of message reception and cyclostationary features, highlighting the method's ability to conceal messages from hostile receivers. Challenges encountered are discussed, with suggestions for future enhancements to improve real-world applicability.

Doctor of Philosophy, The Ohio State University, 2024, Consumer Sciences

Risk preference plays a central role in determining the financial and social well-being of individuals and households. In the three studies of this dissertation, I explore whether individuals' risk preferences over wealth and/or health vary systematically with exogenous changes in general education, particular education types, and health status. The analysis is also informed by recent empirical evidence that genetic differences across individuals are associated with socio-economic characteristics and behaviors. In Chapter Two, I explore the association between general education and risk preference over wealth. The study modifies a theoretical framework to introduce a mechanism through which education is associated with risk preference. Results indicate the existence of the association between education and calculated risk preference over wealth, but do not support the hypothesis that the relationship is causal. In Chapter Three, I introduce a more nuanced measure of education, specific majors/types of education, that aligns more closely with the theoretical construct. I employ local labor market statistics as instruments to predict individuals' choice of college major, discovering that students in math-related fields generally report being less risk-averse. Chapter Four transitions to exploring health status as another potential determinant of risk preference. I question whether risk preferences are consistent across domains and how these preferences shift with changes in health status. Using an instrumental variable approach and an over-identification strategy based on genetic variation associated with nine health conditions, I find that individuals with poorer health report being more risk-averse in health-related decisions but less so in other domains. Additionally, I observe that those with deteriorating health are more likely to be a drinker but less likely to be a smoker; for current users, consumption of both goods decreases as their health worse (open full item for complete abstract)

Master of Science, The Ohio State University, 2024, Public Health

Threshold regression, also known as the first hitting time regression model, offers a robust alternative when the proportional hazard assumption of the Cox proportional hazard model is invalid for time-to-event or survival data. This model defines the event time as the first time a latent stochastic process enters a boundary set. When the underlying process follows a Wiener diffusion process, the event time has an inverse Gaussian distribution. Two essential parameters that determine the behavior of the stochastic process are the level at time zero and its mean rate of change. In medical contexts, this models health trajectory with separate functions for initial health status and the mean rate of change for a patient's health trajectory. This flexibility can provide deeper insights for patients' health process. For example, by separately analyzing initial health status and the rate of health degradation, we can identify specific factors or interventions that significantly impact either the baseline health or its rate of change. However, this led to challenges in variable selection, as we must determine which parameters should be included in the regression function for either the initial health status, the degradation rate, or both. This thesis evaluated various variable selection methods under different conditions through both simulation studies and empirical data analysis. The approaches included frequentist methods such as forward selection, backward selection, and broken adaptive ridge threshold regression estimator (ThregBAR), as well as Bayesian methods including the Bayesian Horseshoe and Bayesian LASSO. The scenarios considered were the event time follows an inverse Gaussian Distribution, as the assumptions stated, and where the number of covariates exceeds the number of observations. The methods were compared based on several criteria, including the false positive rate, false negative rate, and true model rate. In general, we found that Bayesian methods h (open full item for complete abstract)