Texas A&M University-Kingsville: AKM Digital Repository
Not a member yet
1689 research outputs found
Sort by
Multichannel seismic velocity analysis of southern Cascadia Margin using seismic traveltime tomography
No full textThe sedimentation and tectonic history of the Cascadia Margin is complex and poorly understood. Using multichannel seismic data from margin-perpendicular Line 6 collected by the U.S Geological Survey (USGS) off Cape Blanco, Oregon, in 1994, we obtained a better image of seismic structures and accretionary wedge sediments. Our results will help understand local deformation within the Cascadia basin and accreted wedge sediments. Seismic data processing up to migration revealed prominent sedimentary structures. Multiple slope failure deposits were observed in the Cascadia basin sediments, which could be traced back to a large slope failure feature observed 8 km landward of the deformation front. A discontinuous bottom simulating reflector (BSR) is observed in the accreted sediments, and is absent beneath the slope failure feature. Furthermore, a large seafloor scarp of ~150 m is observed within the slope failure region, and appears to coincide with a deep-seated fault that potentially acts like a gas migrating conduit. These results suggest a link between the slope failure events observed in this region and the presence of gas hydrates. Furthermore, traveltime inversion was performed on the landward portion of Line 6 to obtain tomographic velocities. The final tomographic velocity model resolves structures down to 1 km below the seafloor, and compares reasonably well with the nearby sonic velocities. This project highlights the need for higher resolution imagery to be able to detect evidences of the interplay of gas hydrates and sediment along a complex margin
The perceptions of teachers and administrators on PBIS pre- and post- implementation
No full textSince 1980, there has been a need for better selection, execution, and substantiation of positive behavioral interventions with behavior conditions (Gresham, 1991; Sugai & Horner, 1999; Walker et al., 1995) and behavior in general. Positive Behavioral Interventions and Supports (PBIS) has been outlined, explained, and reviewed since its inception in the reauthorization of the Individuals with Disabilities Act (1997). This basic qualitative study sought to interpret the influence of teacher and administrator perceptions of Positive Behavioral and Intervention Support (PBIS) pre- and post-implementation based on participant experiences by looking at student outcomes.
This study was deliberate in examining and evaluating the impact of teacher and administrator perceptions on Positive Behavioral Intervention and Supports. Additionally, this study examined the impact those perceptions have on student conduct, academic achievement, and school culture. This study examined campus policies and expectations, school safety, campus interactions, and school reaction to favorable behavior. This research occurred at a rural elementary school in South Texas. This research utilized a qualitative approach to collect data for development of a conclusion about the impact of PBIS. Qualitative data was acquired through in-depth interviews with the teachers and administrators at a rural elementary school in South Texas. Qualitative data was examined to determine the impact of teacher and administrator perceptions of PBIS on student conduct, educational achievement, and school culture within the classroom and campus environment.
The overall findings of this research indicated that teacher and administrator perceptions on PBIS pre- and-post implementation in regard to student behavior, academic achievement, and overall school culture have a significant positive impact. There was no significant difference between teacher and administrator perceptions pre or post implementation in regard to student conduct, academic successes, or school culture
Teachers' and principals' perceptions of factors that contribute to the success of Blue Ribbon schools
No full textCreating and sustaining an organizational culture that contributes to high academic achievement is frequently a problem in schools. This basic qualitative study explored measures implemented in thriving campus cultures and effective leadership on those campuses by
interviewing principals and teachers in Blue Ribbon Schools. The study focused on principals’ and leaders’ perceptions of three different campuses in South Texas. I selected schools that have been awarded Blue Ribbon campus designation in 2019 to explore factors that contributed to these campuses’ success. Additionally, I utilized a combination of research methods and
techniques such as one-on-one interviews, transcribed notes, audio recordings, and artifacts from those involved in the school’s success.
This study aimed to explore factors contributing to students' academic success in a positive school culture in which effective leadership was present. I applied a theoretical framework to help explain the narrative drawn from participants. Several conclusions emerged from this study. In all cases, the Blue Ribbon School process strengthened leadership effectiveness, principals and teachers collegiality, collaboration and teamwork, and positive school culture. These attributes were closely analogous to discussion of leadership effectiveness
and positive school culture, which were reinforced by retelling participants' narratives. The process stimulated an integrated, holistic view of the campuses, particularly from team members who actively contributed and participated in preparing the application. For future research I recommend Blue Ribbon School studies may be carried out using qualitative methodology with quantitative assessment data, in part, to research the possible cause and effect relationships among the varying factors identified in this study and reporting the
school outcomes
Resource selection, habitat influences on population performance, and body size trade-offs of cervids in a nutritionally variable environment
No full textAnthropogenic influences on the landscape are an ever-evolving challenge. Habitat fragmentation can influence the performance, sustainability, and spatial ecology of wildlife populations. Conversion of rangeland to row-crop farming is the most wide-spread landscape alteration in the United States. Nutritional ecology of wild herbivores is inherently tied to the landscape and interrelated to population performance. Agricultural encroachment could create spatio-temporal nutrient fluctuations, thereby inducing periods of enhanced or diminished forage. The nutrient limitation responsible for variable body condition and reproductive trade-offs is largely unknown.
We sought to understand cropland influences on cervid space use and population performance. Further, we evaluated nutritive ties towards energetic allocation of growth and reproduction using captive cervids to understand relationships in wild settings. We accumulated multi-year data on mule deer (Odocoileus hemionus) movement and morphology in the Texas Panhandle to evaluate resource selection and how habitat use influences morphology. In addition, we raised captive white-tailed deer (Odocoileus virginianus) cohorts to maturity where individuals were split into 2 treatments of diets varying in digestible energy. Mule deer exhibited a mild selection of summer crops and strongly selected cropland during the winter to facilitate nutritional intake when rangeland conditions were poor. We identified a threshold where cropland use decreased with > 20-30% cropland density at multiple scales. Cropland use within specific mule deer annual life-history stages enhanced body condition, body mass, and antler size in males. Further, female mule deer that used cropland had increased reproductive success. Captive adult white-tailed deer exhibited lower mass, body condition, and antler size on a low energy diet as opposed to a standard diet. Birthing rates, litter size and fawn birth mass for reproductive females, however, did not differ between treatments.
In summary, conversion of rangeland to row-crop farming may alter nutritional health relationships in future wild cervid populations. Cervids exhibit a physiological cue to reduce growth to lower energetic requirements to continue reproducing. Therefore, as the agriculture-rangeland mosaic continues to change, populations may undergo shifts in morphology to match their nutritional environment. Furthermore, cropland has the capacity to enhance population measures beyond current adaptations, however, greater agricultural densities may inhibit population spread. Our study demonstrates important population level interactions with the nutritional environment for prominent large mammal species to aid continued management in the foreseeable future
Parallel algorithms for community detection using distributed memory systems
No full textEntities in many real-world networks organize themselves into groups based on relationships between them. Such grouping or clustering forms natural communities, whose entities are well related to other entities in the same community. Through this thesis work, the problem of finding such communities is investigated. This problem, known as community detection, helps understand the structural properties of many networks and is used in many disciplines including social sciences, communication networks, physical sciences, and many more.
Designing efficient and practical algorithms for community detection is a well-studied problem in many disciplines. These disciplines have defined communities in different ways that are relevant to their specific context. Some approaches to community detection are simple and efficient, like label propagation algorithm (LPA), while others are more time-consuming like the Louvain method.
In the current work, a new approach called vector label propagation (VLP) that is based on the label propagation algorithm (LPA) is proposed. This new algorithm is faster and simpler than previous algorithms like clustering algorithms and optimization algorithms based on goodness measures like modularity. It is faster on real-world networks like ego-Facebook (3.2 times) and wiki-vote (5.6 times) than the modularity-based Louvain method. This algorithm also addresses some problems associated with LPA like oscillating labels and instability of solutions. This thesis also provides a variation of VLP that uses weights based on common neighbors in the process of assigning communities. Parallel algorithms using distributed memory systems are presented for both the label propagation algorithm (LPA) and the vector label propagation (VLP) algorithm. A hybrid version using shared memory threads is also proposed which improves scalability up to 192 processors for large graphs
A novel hybrid secure edge server architecture and CDN to enhance the security and response time for edge devices
No full textThe Internet of Things (IoT) is grabbing the attention of scientists all over the globe. With its explosive growth along with applications that require real-time computing power has emerged a new technology called Edge Computing. As a result, Edge Computing changed the way data is processed and handled back and forth for millions of devices around the world, such as autonomous vehicles and electric cars. The confinement of the Cloud Computing technology such as a Content Delivery Network (CDN), contributed significantly to the development of Edge
Computing. Nowadays, this technology can meet the demands of the ever-growing mobile devices and IoT. This Thesis work provides an innovative framework consisting of a hybrid storage architecture consisting of (CDN), Edge Computing, and centralized storage. The centralized
storage consists of multi-level storage systems that comprise Solid-State Drives (SSDs) and Hard Disk Drives (HDDs), which can provide an optimal data storage solution for a wide variety of realtime data processing applications. Transforming the data back and forth between the SSDs and the HDDs plays a crucial role in achieving high performance while meeting the deadline of the edge device request's. Additionally, a new dynamic solid-state disk partitioning mechanism is introduced to optimize security for the proposed framework, among hard disk drives where a partition from the solid-state disks to hard disk drives is assigned based on the workloads of the hard disk drives
Impact of work progress measurement methods on earned value management analysis
No full textWork progress measurement has an important role in in managing construction projects. It provides information for project planning, monitoring and control. Progress of work in construction projects is a key information in the project management process as its deviations from
the project baselines shed lights on the performance of the project. The progress rate is an important index for timely and accurate decision-making in project control. CII (1987), Thomas & Mathews (1996) and Fleming & Koppleman (1996) defined several work progress measurement methods like, supervisor’s opinion, unit completed method, start-finish method, incremental
milestone method, earned value method, estimate percentage complete method and many more.
Project management has one technique, which combines the data regarding cost and schedule of projects to define their current status. This technique is known as Earned Value Management (EVM). The progress of work is a key input to the EVM. The choice of the progress measurement
method is either indicated in the project plan or is at the discretion of the project manager. For many items of work more than one method can be used to measure the progress of work. In this research the impact of various work progress measurement methods on EVM analysis is
investigated to determine if the choice of the method has a significant effect on the results. I have used two work progress measurement methods by CII (1987), Cost-Ratio, Start-Finish and Incremental Milestone. Which gave detailed progress about a sample project in terms of variances and indices
Toward high securuty and energy saving of IoT in edge computing using heterogenous architecture
No full textThe rapid development of the Internet of Things (IoTs) and machine learning techniques has evoked a new computing model called Edge Computing. To boost the functionality, energy-conserving, and safety of IoTs apparatus in edge computing, I proposed a heterogeneous multiprocessing architecture (Saver) in edge computing. Two processing processors run concurrently where Graphics Processing Unit (GPU) and Digital Signal Processing (DSP) are used. Additionally, I studied the effect of using these hybrid processors on edge computing regarding the workload dimensions, energy efficiency, and electricity consumption. I proposed a business rules engine that was considered an essential block in the entire system. The business rules engine was responsible for classifying and simulating different calculations to extract client request patterns and behavior. The business rules engine was also in charge of managing the GPU interactions and the DSP and allowed the machine to use them simultaneously. The business rules engine helped with comments that help to divide the GPU and DSP workload, leading to higher performance and lower energy consumption. The advantage of this approach was that all cores would be used simultaneously. Additionally, I designed and implemented a Secure Socket Layer (SSL) that opportunistically performs cryptographic operations to DSPs and GPUs. I conducted a comparison experiment using Nvidia GeForce RTX 2060 GPU plus a Texas Instrument C6000 high-performance DSP. Saver will generate the following outcomes: First, the proposed research will make innovative contributions to edge computing research. Second, it will cause considerable performance, bandwidth improvements, lower power consumption, and security for the design of IoTs applications, including wireless storage applications
Economic assessment for establishing a recycling plant: case of South Texas area
No full textRecycling retrieves wastes into usable materials and catalyzes economic security by repurposing depleted resources and creating jobs. Most small-sized cities have no recycling process facility hence, there is no recycling collection program. This research analyzed the cost
and economic impact associated with building and operating a recycling facility in the South Texas area. Utilizing the waste input-output model it was discovered that at a recycling rate of 28%, about 4000 tons of recyclables processed, the revenue generated each year is slightly above 250,000 while
creating 36 new employment opportunities. The cost-benefit analysis outputs indicated a negative net present value at a 4% discount rate over 10 years. Although the outputs show a negative net worth, this research suggests establishing the facility with government support in
grants and tax incentives to shore up the profit margin of the recycling plant. This research also suggests the location of the facility using a nonlinear programming optimization model
Verification of a single port ram using UVM methodology
No full textIn this thesis, a Single Port Random Access Memory (SPRAM) was designed and functionally verified by developing a verification environment in the most reusable, scalable and maintainable way. The methodology used in this work to develop a verification environment was the Universal Verification Methodology (UVM), which is an open source System Verilog class based library package that helps in creating hierarchical testbenches and reusable verification components. UVM can be used to generate and organize testbench in a systematic way. It separates stimulus generation from the testbench, which allows verification engineers to reuse the same testbench for multiple designs. It is cross-compatible with multiple vendors, so it can be used across projects in different companies. The purpose of this research was to test the best features of UVM and develop a reusable verification environment. The Design Under Test (DUT) for this study was a 4-bit Single Port RAM that can perform a single read or write operation at any time. Constraint randomization techniques from System Verilog and UVM were used to generate random signals and apply stimulus to the DUT. The results show that the design was functionally correct and bug free. The overall structure of the testbench was defined by extending UVM component and data classes and test cases developed by extending base tests. The simulations were run in EDA playground using Synopsis VCS simulator and UVM v1.2