AUETD (Auburn University)
Not a member yet
9771 research outputs found
Sort by
The Power of Hope: Posttraumatic Growth in Former Partners of the Sexually Addicted Individuals
Get PDFThe purpose of this interpretive phenomenological analysis is to explore and gain an in-depth understanding of the transformative changes emerging from lived experience of former partners of the sex addicts. This study informs the phenomenon of personal growth resulting from individual responses to betrayal trauma. Semi-structured open-ended interviews with a purposeful sample of 12 participants generated data that evidencing that such a seismic event - like the betrayal caused by a sex addict- can be an opportunity for a metamorphosis of a partner’s schemas in the following areas: seeing new possibilities, changed relationships, the paradoxical view of being both stronger yet more vulnerable, a greater appreciation for life, and changes in the spiritual and existential domains (Calhoun & Tedeschi, 1999). The following themes were identified in the current study: (1) changed perception of sense of self; (2) learning to listen and honor one’s body’s signals, (3) new-found personal strength that allowed for facing fears and taking charge of one’s life, (4) journey from spiritual bypassing to authentic spirituality, (5) changes in relating to others, and (6) existential wrestling that led to shifts in worldview
Using Functional Assessment and Mapping Tools to Evaluate Headwater Slope Wetlands in Coastal Alabama
Get PDFHeadwater slope wetlands are a ubiquitous forested wetland type located at the headwaters of coastal streams in the southeastern U.S. Coastal Plain. There is concern that past and current coastal land use / land cover (LULC) change may reduce the capacity for these wetlands to provide important functions (e.g., habitat, water quality improvement, and flood attenuation). To investigate this, 74 headwater wetlands in coastal Alabama (i.e., Mobile and Baldwin County) were assessed for important functional attributes (forest structure, soils, and hydrology) represented by various ecological measures. These data were compared to LULC data (i.e., % forest, urban and agriculture) from each wetland’s catchment over a range of surrounding landscapes typical of the Alabama coast. Wetland attributes were measured using a regionally specific rapid assessment model, the Hydrogeomorphic Approach (HGM) for the functional assessment of headwater slope wetlands in the coastal plain region of Mississippi and Alabama. Significant relationships between wetland shrub cover and agricultural and urban land use suggests LULC change may increase midstories densities. Urban land use was additionally related to increased herbaceous understory coverage and soil dewatering, as well as reductions in soil organic matter content. Despite some significant relationships and notable trends, urban and agriculture were not highly correlated with several other field measurements, suggesting other landscape factors are important for determining the functional capacity of these wetlands. Headwater wetlands can be difficult to map because of their tendency to transition gradually into uplands on the landscape. For the second part of this study, we evaluated the Wetland Intrinsic Potential (WIP) tool and its use of multi-scale topographic indices, hydrologic proxies, and random forest procedures that contribute to ‘cryptic’ wetland detection in the Bushy Creek – Dyas Creek watershed, near Bay Minette, Alabama. An initial model was trained and validated on a spatial subset of the watershed to predict headwater wetland presence, absence, and extent. The model was then applied to the remaining spatial extent of the watershed. Overall accuracy for the secondary validation dataset was 92.3%, with wetland omission and commission errors of 14.0% and 4.5%, respectively. Our statistical analyses indicated WIP reliably discerned wetlands from uplands. These findings can be used to infer the applicability and limitations of this method for wetland mapping along the northern Gulf of Mexico and support future models which explore land use/cover and hydrogeomorphic relationships with wetlands. Ultimately, information gained from this thesis study will assist in wetland monitoring efforts to better assess the environmental services provided by coastal drainages along the Gulf coast
Development of Hybrid Model for Estimating Liquid Entrainment Fraction and Uncertainty: Integrating Machine Learning and First Principles
Get PDFIn two-phase annular flow, a thin liquid film forms around the pipe as gas flows through the pipeline center. Some portion of the liquid is carried into the gas core. The liquid entrainment fraction, which is defined as the flow rate of the entrained liquid droplets relative to the overall liquid flow rate, is an essential parameter for accurately estimating pressure drop, flow rate, liquid holdup, and dry-out conditions in annular flow. Accurate estimates of these variables are essential for the design and operation of wet gas transmission pipelines, pipeline corrosion inhibition wellbore and flow line design, and downstream separation design and optimization. Numerous first-principle models exist for predicting liquid entrainment fraction in a two-phase flow. However, due to the intricate complexity of the entrainment phenomena, none of the models incorporates effects from all observations nor can they be extended across a wide range of operating conditions, which results in inaccurate estimation of the liquid entrainment fraction. Moreover, none of these models is developed with the capability of quantifying the entrainment fraction prediction uncertainty.
In this dissertation, a hybrid modeling framework combines first-principle model and data-driven model is developed to estimate the liquid entrainment fraction with its uncertainty in two-phase flow. A database composed of 1,662 entrainment fraction experimental measurements is used for predicting the liquid entrainment fraction for three different flow orientations. The first-principle model predict the liquid entrainment fraction while the data-driven model predict the model discrepancy, which is defined as the difference between the experimental measurements and the first-principle model prediction. Different machine learning techniques and uncertainty quantification methodologies are applied to estimate entrainment fraction with its uncertainty within the hybrid model. To pick the best model, a novel metrics for evaluating the performance of machine learning model with stochastic output called uncertainty width is proposed to compare the model performance and best model is picked for each flow orientation. As a result, Bagging Gaussian Process Modeling (GPM) with estimated noise is identified as the best model with the best accuracy and uncertainty calibration.
The hybrid model performance is enhanced using different methodologies. To extend the hybrid model’s applicability from laboratory to field scale, dimensional analysis (DA) is performed to obtain dimensionless numbers used as the updated inputs. To prune the irrelevant inputs, a novel Gaussian Process embedded feature selection approach called Derivative Decomposition Ratio (DDR) is proposed and its performance is compared with another feature selection approach derived from normalizing the sensitivity. To extend the hybrid model’s capability to assistant the development of mechanism, a partial derivative based framework delivering quantitative first-principle model refinement decisions is proposed in this study.
The methodology developed in this dissertation can be applied to estimate the liquid entrainment fraction with its uncertainty for three flow orientations given operating conditions. The methodologies developed in this study, such as the feature selection method and the first-principle model refinement decisions, can also be used in other applications in addition to the liquid entrainment fraction. Those methodologies can be considered for any GPM feature selection and any mechanism refinement studies
Cover Crop Management and Nutrient Type Effects on Collard and Cotton Growth
No full textProducing fresh vegetables and fiber, such as collards (Brassica oleracea L. var. viridis) and cotton (Gossypium herbaceum L.), in cover crop residue while minimizing soil disturbance can be helpful in lowering environmental impacts. Furthermore, using different types of fertilization, including poultry litter, for cotton can also promote soil health and manage input costs. Two experiments were performed to investigate 3 different residue management methods including rolled/crimped (Roll), mowed (Mow), and mowed + incorporated by tillage (MowIncorp) and their effects on a collard and cotton production system.
Collards were grown in both an iron clay pea (Vigna unguiculata L.) and pearl millet (Pennisetum glaucum L.) cover crop and results showed that cover crop type influenced collard productivity in the 2015, 2016, and 2017 seasons. Pearl millet produced more biomass (P=0.0013) with a three-year average of 8461 kg ha−1 compared to 6465 kg ha−1 for iron clay pea. Carbon sequestration was evident in both cropping systems with 10.5% (from 5.04 to 5.57 g C kg-1) and 8.1% (from 5.08 to 5.49 g C kg-1) increases in total soil carbon for pearl millet and iron clay pea, respectively, in the top 15 cm of soil over three seasons. Overall, collard yield in the iron clay pea cover crop produced more (P<0.0001) with a 3-year average of 7268 kg ha-1 compared to 4724 kg ha-1 for the pearl millet.
The cotton experiment investigated the residue management methods for a single rye cover crop and four different nutrient treatments for the 2022-2023 seasons. All nutrient treatments included the same amount of nitrogen (100.9 kg N ha-1) and consisted of: 100% poultry litter at preplant (PL), 67% poultry litter at preplant + 33% liquid urea ammonium nitrate (UAN) at sidedress (PLUAN), 33% granular fertilizer pre-plant + 67% liquid urea ammonium nitrate at sidedress (FERT), and a no nutrient applied control (NONE). Sidedress applications were performed prior to flowering using a coulter with knife spaced 10 cm from cotton row. Results show that nutrient treatments were the main influence on cotton growth with 11% taller plants, 31.8% greater dry weight, and 13.4% more cotton bolls, respectively, for the inorganic fertilizer treatments (FERT) compared to the poultry litter (PL) treatment in 2022. The FERT treatment had 10.7% taller plants, 16.9% greater dry weight, and 9.3% more cotton bolls compared to the PL treatment in 2023. The Mow method hindered emergence for the 2022 and 2024 seasons where rye dry biomass was 6464 and 5982 kg ha-1, respectively, but no differences in emergence were evident across methods in 2023 with rye biomass production of 3723 kg ha-1. For both seasons, no significant differences were realized between any of the nutrient treatments for seed cotton yield, but the yield was lower for MowIncorp in 2023 compared to the Roll and Mow. In conclusion, poultry litter was successful in sustaining cotton yield similar to inorganic fertilizers under cover crop residue management without tillage
Rapid Detection and Accurate Discrimination of Microorganisms by Liquid Chromatography, High-Resolution Ion Mobility, and Tandem Mass Spectrometry
Get PDFDetermining bacterial identity at the strain level is crucial for public health to enable appropriate medical treatment and reduce hospitalization times and antibiotic resistance. To achieve this goal, we have reported the coupling of ambient ionization techniques with a commercial drift tube ion mobility mass spectrometer and demonstrated their ability to rapidly separate constitutional and geometric isomers. After successful rapid isomer separation, we investigated paper spray – IM – MS/MS to discriminate five Bacillus species. We optimized several parameters, such as the incubation time and the spray solvent. We found that a 4 h-incubation time is sufficient for detection and that isopropyl alcohol (IPA) gives a longer spray time and higher intensities of the observed biomarkers than methanol (MeOH, typical spray solvent in PS – MS experiments). Numerical multivariate statistics (principal component analysis followed by linear discriminant analysis) allowed discrimination at the species level with a prediction rate of 92.4 % and 97.6 % using the negative and positive ion information from PS – MS/MS, respectively. However, when including the corresponding drift times of the biomarkers, i.e. PS – IM – MS/MS, prediction rates of 99.7% and 100% were obtained using the negative and positive ion information, respectively. We attribute the improvement in prediction rates to the ability of IM separations to resolve isomers.
When analyzing seven E. coli strains by PS – IM – MS/MS, the prediction rate was 80.5% after numerical data fusion of negative and positive ion modes. Therefore, we combined liquid chromatography with IM – MS/MS as LC – IM - MS/MS to accurately discriminate the seven E. coli strains. Numerical multivariate statistics demonstrated the ability of this method to perform strain-level discrimination with prediction rates of 96.1% and 100% using the negative and positive ion information, respectively. This work demonstrates the great potential to accurately detect pathogenic and antibiotic-resistant bacteria in agrochemical screening, disease diagnostics, etc. Moreover, this work can pave the way for developing standalone ion mobility spectrometers which can be used in several medical, environmental, and security applications
AI-aided System and Design Technology Co-optimization Methodology Towards Designing Energy-efficient and High-performance AI Accelerators
Get PDFThe rapid growth of artificial intelligence (AI) and deep learning (DL) workloads has created an urgent need for more efficient and high-performance AI accelerators, both at the edge and in cloud data centers. The computational and memory demands of large models, such as ChatGPT and Sora, have far outpaced advancements in semiconductor technology, leading to the emergence of the memory wall and area wall. These challenges necessitate the exploration of new technologies and methodologies. This dissertation presents a comprehensive investigation into emerging memory technologies, innovative architectural designs, and optimization methodologies aimed at improving energy efficiency, performance, and area utilization in AI accelerators.
First, we introduce a high-performance AI accelerator that incorporates spin transfer torque magnetic RAM (STT-MRAM) as the on-chip memory system. Through model-driven design space exploration, we develop a novel scratchpad-assisted buffer architecture that
optimizes memory retention time, read/write latency, and energy efficiency by dynamically adjusting for process and temperature variations. Our STT-MRAM-based design (STT-AI) achieves a 75% reduction in area and 3% power savings compared to SRAM-based systems, with minimal trade-offs in accuracy, demonstrating its suitability for modern AI workloads.
Next, we address the limitations of existing accelerators in handling large-batch AI training and inference due to memory bandwidth and capacity constraints. We propose a design technology co-optimization (DTCO)-enabled memory system utilizing spin-orbit torque magnetic RAM (SOT-MRAM) to significantly increase on-chip memory capacity. The limitations posed by STT-MRAM are also addressed by introducing SOT-MRAM. This workload-aware memory system shifts AI accelerators from being memory-bound to achieving system-level peak performance. Our results show an 8× improvement in energy efficiency and 9× reduction in latency for computer vision benchmarks, along with substantial gains in natural language processing tasks, while consuming just 50% of the area compared to SRAM at the same capacity.
Finally, to address the limitations of large monolithic designs, we explore the potential of chiplet-based architectures for AI accelerators. The vast design space and complex trade-offs between power, performance, area, and cost (PPAC) require a systematic optimization approach. We introduce an optimization framework, Chiplet-Gym, which integrates heuristic-based methods, such as simulated annealing (SA), with learning-based algorithms, such as reinforcement learning (RL), to evaluate and optimize chiplet-based AI accelerator designs by accounting for resource allocation, placement, and packaging architecture. Our results indicate that reinforcement learning demonstrates greater stability and achieves a 16% higher cost model value than simulated annealing. The framework-suggested design
choice delivers a 1.52×improvement in throughput, a 0.27×reduction in energy, and a 0.89×lower cost compared to monolithic designs at iso-area, underscoring the potential of chiplet architectures for the next-generation AI hardware
Role of Solvent Coordination and Ligand Structure on the Electrochemical Redox Cycle of Nickel (II) bis(dithiocarbamate) Complexes for Energy Storage Applications
No full textNickel (II) diethyldithiocarbamate, NiII(Et2dtc)2, undergoes a 2e- ligand-coupled electron transfer (LCET) oxidation at a single potential to form [NiIV(Et2dtc)3]+ in acetonitrile (MeCN) solvent. However, in different nonaqueous solvents NiII(dtc)2 can exhibit divergent redox behavior based on the solvent’s coordination properties, producing significant changes on the electrochemistry of the complex during oxidation and reduction. Pyridine (Py), dimethyl sulfoxide (DMSO), dimethyl formamide (DMF), MeCN, methanol (MeOH), acetone (Ac), and dichloromethane (DCM) were studied here due to their dissimilar coordination abilities towards the nickel metal center. Cyclic voltammetry data in the above-mentioned solvents show distinct behavior which is expected as solvent coordination ability varies. For low coordinating solvents like MeCN and Ac, 2e- oxidation of NiII(Et2dtc)2 to [NiIV(Et2dtc)3]+ occurred at a single potential. Stronger coordinating solvents like MeOH, DMF, DMSO, and Py disrupted the 2e- oxidation by coordinating to the Ni(III) intermediate to form [NiIII(Et2dtc)2(sol)x]+ complexes. The decay of these complexes through ligand exchange with NiII(Et2dtc)2 and disproportionation to yield [NiIV(Et2dtc)3]+ was monitored as a function of scan rate and temperature to extract rate constants and activation parameters. A thorough analysis of activation parameters revealed that ΔHapp‡ generally increased with solvent donor number, suggesting solvent dissociation was a key factor in the rate limiting step. However, ΔSapp‡ was found to be negative for all solvents, suggesting an associative mechanism in line with dimer formation with NiII(dtc)2 to facilitate ligand exchange. In a mixed solvent composition of acetonitrile, pyridine, and dichloromethane in a ratio of 90:7.5:2.5, the apparent activation entropy (∆Sapp‡) decreases as the steric bulk of the substituent R groups replacing an ethyl (Et) group increase. This trend suggests that greater steric bulk promotes a more associative mechanism. Primary amine-based dtc ligands were introduced to improve solubility of Ni(II) complexes where solubility of NiII(Etdtc)2 was four-fold than NiII(Et2dtc)2 in MeCN
A Comprehensive Study of Yield Components, Nutrient Uptake, Root Characteristics, and Cover Crops for Alabama Cotton Production
No full textThis study has yielded crucial findings across multiple aspects of cotton agronomy and breeding by systematically evaluating 20 upland cotton (Gossypium hirsutum L.) cultivars released over the past 65 years (1953-2018). The first component of the research investigated the biomass partitioning and yield performance of the cotton cultivars. The results revealed a substantial increase in lint yield, averaging 11.7 kg ha⁻¹ yr⁻¹ in 2020 and 12.6 kg ha⁻¹ yr⁻¹ in 2021, driven primarily by improvements in lint harvest index and total aboveground biomass. However, the study cautions that merely increasing total biomass without concomitant enhancements in reproductive partitioning could lead to decreased lint yield, as total biomass had a direct negative effect on lint yield. The findings underscore the importance of prioritizing improvements in the efficiency of partitioning aboveground biomass into the economically valuable lint fraction to achieve further yield gains.
Alongside the aboveground traits, the research also explored the evolution of nutrient dynamics in these twenty cotton cultivars. Over the 65-year period, the study observed a steady increase in total nutrient uptake (N, P, and K) and internal nutrient use efficiency. While nutrient concentrations in the vegetative growth either decreased or remained unchanged, there was an increase in nutrient concentration in the seed. This shift in nutrient partitioning suggests an indirect improvement in allocating essential minerals to the economically valuable seed component. Averaged over all tested nutrients, modern cultivars released in 2018 were 27% more efficient in producing lint per unit of nutrient uptake than the 1950s, highlighting the potential for further advancements in nutrient management strategies.
iii
Complementing the aboveground analyses, the research also delved into the root traits of the cotton cultivars. The investigation revealed significant cultivar differences in carbon isotope discrimination (Δ13C), a proxy for water use efficiency, and variations in root architectural characteristics, such as total root length, root surface area, and root crown attributes. Interestingly, the total root crown surface area exhibited a linear increase with the year of cultivar release, suggesting that breeding efforts to enhance aboveground performance have indirectly improved certain root parameters. Furthermore, the distribution of root surface area across different soil depths showed changes, with modern cultivars allocating a higher proportion of roots in the topsoil compared to older cultivars. These findings underscore the untapped potential for improving cotton's root system architecture and water use efficiency through targeted breeding strategies, which could contribute to developing climate-resilient cultivars.
In addition to the cultivar evaluation, the research also explored methodological advancements in root phenotyping. Specifically, the study investigated the influence of X-ray computed tomography (CT) system voltage on the penetration capability in diverse soil types and container sizes. The results demonstrated that increasing the voltage enhances image quality up to a certain plateau, providing valuable guidance for researchers utilizing X-ray CT to study root growth and development in complex soil environments.
Finally, evaluating six years of continuous cover crop systems in the southeastern United States provided insights into their effects on soil organic carbon and cash crop performance. While the results varied across sites and years, adopting high-biomass cover crop species, such as rye and crimson clover, showed the potential to improve soil carbon content and enhance cash crop yields, particularly under irrigated conditions. These findings underscore the importance of continued
iv
research and the development of region-specific recommendations for integrating cover crops into existing cotton production systems.
Collectively, this comprehensive research program has explained the key drivers of historical improvements in cotton lint yield, the evolving nutrient dynamics of modern cultivars, the untapped potential of belowground traits, the optimization of imaging techniques, and the impacts of cover cropping. These insights can inform the strategic development of future cotton cultivars and management practices, ultimately contributing to the sustainable intensification of cotton production to meet the global demand for this vital fiber crop
AI Innovations: Ensemble Learning and Energy-Efficient Knowledge Injection
No full textThis research expands on the recent rapid advancements in Artificial Intelligence (AI) by focusing on its practical applications and environmental impact. It introduces a novel ensemble learning system for Network Intrusion Detection that integrates a one-dimensional Convolutional Neural Network (1D-CNN), an FT-Transformer, and XGBoost. It leverages the spatial pattern recognition of the 1D-CNN, the self-attention capabilities of the FT-Transformer, and the efficiency of XGBoost to significantly enhance detection accuracy, precision, recall, and F1-score compared to existing systems. Toward heart disease prediction, this research develops an ensemble model that combines BERT, FT-Transformer, and XGBoost, which excels at extracting meaningful features, capturing temporal patterns, and handling structured data. It improves diagnostic accuracy and has the potential to revolutionize healthcare diagnostics. In addition, this research examines the environmental impact of Large Language Models (LLMs), focusing on their energy consumption during the knowledge-injection process. When using knowledge injection, we compare the energy efficiency of Fine-tuning and Retrieval-Augmented Generation methods. The findings underscore the importance of enhancing AI's practical capabilities while mitigating its carbon footprint
Are you Satisfied? A Look at How Adult Attachment Style and Perfectionism Influence Romantic Relationship Satisfaction
Get PDFThis study is important in expanding the existing literature on the Perfectionism Social Disconnection Model (PSDM) and building an understanding of the connections between adult attachment style, maladaptive perfectionism, and romantic relationship satisfaction. This cross-sectional study examined the relationships between attachment style, maladaptive perfectionism, and romantic relationship satisfaction among a convenience sample of 214 adults currently in a romantic relationship. Based on previous literature and theoretical models, this study hypothesized that insecure adult attachment styles (anxious and avoidant) would directly and indirectly predict romantic relationship satisfaction for individuals currently in a romantic relationship through three mediation pathways of rigid perfectionism, self-critical perfectionism, and narcissistic perfectionism. Path analysis was conducted to test the hypothesized model. The current study found a direct link between higher levels of attachment avoidance and lower romantic relationship satisfaction. There was not a significant relationship between attachment anxiety and romantic relationship satisfaction. Further, there were direct links between insecure attachment styles (avoidant and anxious) and maladaptive perfectionism. Unexpectedly, there were no direct links between maladaptive perfectionism and relationship satisfaction. Participant’s sex and current relationship length were used as control variables. Implications for future research include the necessity for further exploration of the interpersonal impacts of insecure attachment styles and maladaptive perfectionism. Perhaps incorporating more narrow interpersonal concepts (e.g., relationship length) instead of broad interpersonal concepts (e.g., relationship satisfaction) may allow for successful expansion of the PSDM. For practitioners, this research can help identify appropriate interventions to use with individual clients or couple’s therapy and/or which areas to provide psychoeducation on with their clients