The LAIR at East Texas A&M
Not a member yet
7179 research outputs found
Sort by
Graphical Processing Unit Accelerated RNA Substructure Comparison and Search Engine
No full textExponentially increasing Ribonucleic acid (RNA) secondary structure databases have presented new challenges to researchers in the field and motivated the idea for fast preparation of collection of huge RNA structure database for subsequent analysis. Further when big data discussion happens, consideration of the processing time is a must. Use of big data analytics and Graphics Processing Unit (GPU) accelerated deep learning have proven highly beneficial in computationally extensive data analysis which motivated this study to incorporate GPU to accelerate the existing as well as new comparison and search algorithms.The recently developed comparison and search algorithm yielded efficient results with the use of newly proposed relative addressing based (RAB) RNA secondary structure representation. The RAB representation embeds the 2D structure information of an RNA into a sequence. It allows to store the RNA structure database into a suffix array and further assists the development of fast substring search and comparison algorithms. The algorithms were tested on databases of around 5000 RNAs. Now as the database sizes have reached collectively to millions, while performing analysis, limitations have been observed with existing algorithms when used over huge databases. Hence, the goals of this project include automation of the steps to streamline the handling of RNA structure data curation such that the database could be refreshed as needed in accordance with its constantly changing nature. Some new search and comparison problems have been formulated for RNA substructures analysis. These problems will show meaningful impact on RNA structure analysis, done on large databases. The development of algorithms to solve these problems efficiently will be the focus of this study. The efficiency of proposed solutions would be proved by showing comparative test results. In addition, the practical use of this study in life-sciences will also be discussed. Finding similarity by comparing RNA structure sequences is substantial for structural bioinformatics researchers but upon comparison, the computational cost could outweigh the gain (Stern & Mathews, 2013). Hence, this study proposes to give improved solution for the previously developed RNA substring search and comparison problems and for new discussed problem scenarios to work efficiently on large databases with a Compute Unified Device Architecture (CUDA)-supported General Purpose Graphics Processing Unit (GPGPU) programming. All the application development would be done using Python language, which is popularly used nowadays for big data analytics and machine learning. The comparative test results would be generated on system with NVIDIA GPU support
A Deep Learning Approach for Improving Activity Detection on Portable Smart Devices
No full textActivity trackers have become very popular in the recent decade. Human activity recognition (HAR) has been receiving more attention for years because of its easy-to-use features in numerous applications to identify activities. Activity tracker users want to measure their activity for health, fitness, or similar purposes. Trackers are loaded with sensors and researchers are aimed at increasing software functionality of these devices. Furthering simple tasks already presented in many trackers, such as number of steps and number of calories burned, we developed pattern recognition software to be used with the activity trackers. We advanced a small-size artificial intelligent agent to detect user activity. Our classification model can be deployed on activity tracker sensors-based devices, such as phones, smart watches, and smart activity trackers. Our primary aim is to recognize four human activities: walking, walking upstairs, walking downstairs, and standing. We used two deep learning methods. They are Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) artificial recurrent neural network to learn users’ activity characteristics and predict activity. For the dataset, we utilized one of the most used open-source datasets of Smartphone –Based Recognition of Human Activities and Postural Transitions, which was collected from 30 individuals who performed postural transitions and basic activities. These were recorded using the smartphone’s sensors. We used TensorFlow as the primary software library with Keras API. The significant finding of this research is the increased detection accuracy by exceeding previously published results. The leave-one-subject-out experiments yielded more than 97% classification accuracies in the recognition of walking, walking upstairs, walking downstairs, and standing. A feasible window size on data and deep learning architecture are determined through the experiments. Our research provides many benefits to people and healthcare workers who would like to accurately measure the four fundamental activities we targeted. Healthcare providers are the primary beneficiary for newly developed software as they may prescribe these activities to patients
Computer-Based Collaborative Multimodal Writing in the French as a Foreign Language Context
No full textIn recent decades, a plethora of empirical work has indicated that collaborative writing (CW) affords more opportunities for collective scaffolding and language negotiation (Li & Kim, 2016; Li & Zhu, 2013; Storch, 2019) and promotes the development of second language (L2) writing skills (Bikowski &Vithanage, 2016; Storch, 2013; Strobl, 2014). Concurrently, previous studies have also demonstrated that multimodal writing (MW) benefits L2 learning as it increases learners’ awareness of metafunctions of semiotic modes (Shin et al., 2020) and overall writing proficiency (Oskoz & Elola, 2016; Vandommele et al., 2017). Although many studies have investigated students’ MW processes (e.g., Hafner, 2015; Shin et al., 2020; Smith et al., 2017), research examining finished multimodal products holistically in various semiotic aspects is lacking. Moreover, as an innovative writing pedagogy, work on collaborative multimodal writing (CMW) is still in its infancy and barely explored, especially in non-English learning settings. To bridge these gaps, this dissertation study sets out to explore computer-mediated MW completed collaboratively in the French as a foreign language (FL) context.Seven elementary-level French FL students worked in three small groups to complete a MW task in which they jointly created digital postcards describing their vacation activities. Using a multiple-case study design, I closely examined the following key areas: (1) the process by which the students collaboratively worked to complete the MW task; (2) the quality of their writing products; and (3) their perceptions. The triangulation of multiple data sources (i.e., audio- and screen- recordings, Google Docs history records, students’ finished multimodal texts, pre-and post-task questionnaires, and semi-structured interviews) revealed several findings. First, the groups made strategic modal choices by drawing on various semiotic resources, including written texts, images/drawings, and hyperlinks, to co-construct their multimodal products. Second, three distant patterns of interaction were identified: Group 1 (Collectively contributing/Mutually supportive), Group 2 (Dominant/withdrawn—Cooperative), and Group 3 (Active/Passive), and the patterns of interaction were found to be connected to the qualities of the writing products. Finally, several themes emerged concerning perceived affordances (e.g., enhanced writing skills, awareness of semiotic resources, and convenient synchronous writing via Google Docs) and constraints (e.g., unequal participation, difficulty with co-ownership, and technical difficulties)
Discovering and Modifying Latent Directions of Synthetic Images while Maintaining Image Realism
No full textIn recent years, Generative Adversarial Networks (GANs) have made immense progress in their ability to generate high quality synthetic images. However, even though these generators can now produce high resolution images, modifying synthetic images whilst also preserving their realism has not developed as rapidly. In this paper, we will be studying the issue of poor synthetic image modifications. We will begin this by looking at the latent space that is sampled to generate the images as well as attempting to identify and modify the latent space of the image generated by a GAN (which changes certain important features of the images). By doing so we can create a set of controls that will let us directly modify specific aspects of the images while keeping the resolution optimal. We will also create an architecture that can generate synthetic images that look like a given input image
La Enseñanza Del Español Para Hablantes De Herencia En Dallas-Fort Worth: Perfiles Profesionales Y Formación Docente
No full textEste estudio exploratorio, inédito en el área de Dallas-Fort Worth (DFW), tiene como objetivo identificar el perfil profesional de los docentes de español como lengua de herencia (LH) para establecer si poseen las capacidades necesarias para enseñar a los estudiantes hablantes de herencia (HH) de la zona, cuyo número ha venido crecido aceleradamente, lo que ha llevado a la creación de clases y currículos especiales para ellos. Para llevar a cabo esta investigación, se realizaron encuestas anónimas y entrevistas dirigidas a docentes de español de escuelas secundarias de los diferentes distritos escolares del área, con el fin de conocer su formación profesional en la enseñanza de la lengua, así como sus creencias y sus perspectivas sobre los estudiantes HH, el currículo que enseñan, el apoyo que reciben de sus distritos escolares, las alternativas de actualización profesional disponibles y el futuro del español en la zona. El análisis de los datos recopilados permite conocer algunas de las dificultades a las que se enfrentan los docentes de HH en esta parte del estado de Texas. El estudio termina con una serie de recomendaciones para mejorar la enseñanza de español como LH en las escuelas del área
An Exploration of Elementary Teacher Perceptions of Student-Centered and Teacher-Centered Instruction
No full textKappa Delta Lion Statue
No full textA color photograph of the Kappa Delta Lion statue. The northeast entrance to the Rayburn Student Center can be seen in the background.https://lair.etamu.edu/scua-univ-photos-browse-all/1359/thumbnail.jp
Keith D. McFarland Science Building Exterior
No full textA color photograph showing the northeast face of the Keith D. McFarland Science Building.https://lair.etamu.edu/scua-univ-photos-browse-all/1363/thumbnail.jp
Pride Rock Residence Hall Exterior
No full textA color photograph showing a main entrance to Pride Rock Residence Hall. A walkway lined with picnic tables leads up to the building.https://lair.etamu.edu/scua-univ-photos-browse-all/1373/thumbnail.jp
Whitley Residence Hall Exterior
No full textA color image showing the north face of Whitley Residence Hall.https://lair.etamu.edu/scua-univ-photos-browse-all/1376/thumbnail.jp