Design Of Learning Model Research Articles

A Framework for Multi-Task Learning in Dynamic Adaptive Streaming Over HTTP

This paper presents a framework with a taxonomy for multi-task learning in the context of Dynamic Adaptive Streaming over HTTP (DASH). DASH is a widely used technology for video streaming, and multi-task learning has emerged as a promising approach to enhance the performance and user experience of DASH systems by jointly optimizing multiple related tasks. The framework provides a structured approach to design, train, and evaluate multi-task learning models in DASH, while the taxonomy categorizes the key components and approaches within the framework. The taxonomy includes task types, multitask learning approaches, input features, and training strategies. Task types encompass video quality adaptation, buffer management, bandwidth estimation, content pre-fetching, and resource allocation, representing the specific tasks involved in DASH. Multi-task learning approaches encompass methodologies such as shared representation learning, task-specific layers, multi-head architectures, knowledge distillation, and reinforcement learning, offering flexibility in model design and optimization. Input features cover video characteristics, network conditions, device capabilities, and user preferences, providing the necessary information for informed decision-making across tasks. Training strategies include joint training, alternate training, hierarchical training, task weighting, and task balancing, determining how the multi-task learning model is trained and optimized. By following the presented framework and taxonomy, researchers and practitioners can systematically approach the design, training, and evaluation of multi-task learning models in DASH. The framework enables the development of efficient and adaptive video streaming systems by leveraging the interdependencies among tasks. The taxonomy helps organize the components and approaches within the framework, aiding in a better understanding of the various aspects of multi-task learning in the DASH context. Overall, this framework and taxonomy provide a valuable resource for advancing the field of multitask learning in the dynamic and complex domain of video streaming over HTTP.

Shape sensing of optical fiber Bragg gratings based on deep learning

Continuum robots in robot-assisted minimally invasive surgeries provide adequate access to target anatomies that are not directly reachable through small incisions. Achieving precise and reliable shape estimation of such snake-like manipulators necessitates an accurate navigation system, that requires no line-of-sight and is immune to electromagnetic noise. Fiber Bragg grating (FBG) shape sensing, particularly eccentric FBG (eFBG), is a promising and cost-effective solution for this task. However, in eFBG sensors, the spectral intensity of the Bragg wavelengths that carries the strain information can be affected by undesired bending-induced phenomena, making standard characterization techniques less suitable for these sensors. We showed in our previous work that a deep learning model has the potential to extract the strain information from the eFBG sensor’s spectrum and accurately predict its shape. In this paper, we conducted a more thorough investigation to find a suitable architectural design of the deep learning model to further increase shape prediction accuracy. We used the Hyperband algorithm to search for optimal hyperparameters in two steps. First, we limited the search space to layer settings of the network, from which, the best-performing configuration was selected. Then, we modified the search space for tuning the training and loss calculation hyperparameters. We also analyzed various data transformations on the network’s input and output variables, as data rescaling can directly influence the model’s performance. Additionally, we performed discriminative training using the Siamese network architecture that employs two convolutional neural networks (CNN) with identical parameters to learn similarity metrics between the spectra of similar target values. The best-performing network architecture among all evaluated configurations can predict the shape of a 30 cm long sensor with a median tip error of 3.11 mm in a curvature range of 1.4 m−1 to 35.3 m−1.

Skin Lesion Classification Using CNN-based Transfer Learning Model

The computer-aided diagnosis (CAD) and the analysis of skin lesions using deep learning models have become common in the last decade. The proposed CAD systems have considered various datasets and deep learning models. The transfer of knowledge from particular pre-trained models to others has also gained importance due to the efficient convergence and superior results. This study presents the design and implementation of a transfer learning model using Convolutional Neural Networks (CNN) with variable training epoch numbers to classify skin lesion images obtained by smartphones. The model is divided into the inner and external CNN models to train and transfer the knowledge, and the preprocessing and data augmentation are not applied. Several experiments are performed to classify cancerous and non-cancerous skin lesions and all skin lesion types provided in the dataset separately. The designed model increased the classification rates by 20% compared to the conventional CNN. The transfer learning model achieved 0.81, 0.88, and 0.86 mean recall, mean specificity, and mean accuracy in detecting cancerous lesions, and 0.83, 0.90, and 0.86 macro recall, macro precision, and macro F1 score in classifying six skin lesions. The obtained results show the efficacy of transfer learning in skin lesion diagnosis.

Machine learning prediction of petty corruption intention among law enforcement officers

Law enforcement agencies face a widespread problem of corruption, which jeopardizes their credibility and institutional integrity. Thus, the primary goal of this study is to develop a machine learning prediction model for petty corruption intentions as an early warning system for law enforcement officials who fail to perform their duties and obligations with integrity. Using a questionnaire survey of two hundred twenty-five participants, from senior officers to rank and file police officers, this study presents the fundamental knowledge on the design and implementation of machine learning model based on six selected algorithms; generalized linear model, fast last margin, decision tree, random forest, gradient boosted trees, and support vector machine. In addition to demographic factors, the efficacy of each machine learning algorithm on petty corruption was evaluated using general strain theory (GST) attributes: financial stress, work stress, leadership pressure, and peer pressure. The findings indicated that peer pressure has given the highest weight of contributions to most of the machine learning algorithms. The most outperformed machine learning in terms of the classification accuracy is gradient boosted trees with accuracy above 90%. This paper presents useful knowledge to enhance the realization of implementing intelligent corruption detection tools.

Sign Language Detection without Sensors

bstract: The time period time device assists communication between the hearing impaired and the general public. We are proud to present the design and implementation of language learning models. In most cases, we create a trust model that distributes signatures regularly. Also, this method is a good way for beginners to use sign language to practice the language. As part of this study, various human-computer interactions for gesture recognition were investigated and analyzed. It has been found that the best way to detect human movement is layered image processing. Even with very little background, the system can detect certain characters in the language with 70-80% accuracy.

Peningkatan Keterampilan Guru dalam Mengembangkan Desain Model Pembelajaran berorientasi HOTS Berbasis Learning Management System bagi Guru Sekolah Dasar di Kota Surakarta

The purpose of this service activity is to improve teacher skills in developing HOTS-oriented learning model designs based on the Learning Management System (LMS) for elementary school teachers in Surakarta City. Participants in the community service activity totaled 30 teachers consisting of 10 class teachers of SD ICT Al-Abidin, 10 class teachers of SD Ta'mirul Islam Surakarta, and 10 class teachers of SD Muhammadiyah 1 Ketelan Surakarta. The method of implementing this service uses training and mentoring. The instrument for evaluating the improvement uses a test and is analyzed descriptively. The results of this service activity show that there is an increase in teacher skills after participating in the training activity. Teacher skills increase in the experience aspect in designing HOTS-oriented learning models, teacher experience in designing activities to develop HOTS, and in processing features contained in the Learning Management System. Based on the results of limited trials, the HOTS-oriented learning model design in the Learning Management System achieves an average student passing scores of 83%.

Model Desain Pembelajaran PPSI Pada Mata Pelajaran Pendidikan Agama Islam

The background of this research is to find out the steps of PAI teachers in the PAI learning design model at SD Negeri 22 Ujung Gurun, Padang City and to find out the strengths and weaknesses in the PAI learning design at SD Negeri 22 Ujung Gurun, Padang City. The method used in this research is qualitative method. The data sources are two informants through interviews, consisting of two Islamic religious education teachers at SD Negeri 22 Ujung Gurun, Padang City. The datas collection techniques are observation, interviews, and documentation. All interview results were then analyzed systematically through four steps of analysis activities, namely (data collection, data reduction, presentation and conclusion). Overall the results of the analysis show; (1) Regarding the steps of the PPSI learning design model at SD Negeri 22 Ujung Gurun, Padang City, there are five: (a) formulating learning objectives (b) Developing evaluation tools (c) Determining teaching and learning activities (d) Planning teaching and learning activities programs ( e) Execution. (2) regarding the excesses and defficiencies of the PPSI learning design model. The results of this study can be used as initial data for further research in examining different issues related to this problem and other issues that are relevant to this context.

Implementasi Model Disain Pembelajaran Dick and Carry dalam Mengembangkan Bahan Ajar Materi Bahasa Indonesia Kelas 3 SDN Kemiri 3

Quality education is the hope and demand of all education stakeholders. In addition, education is an effort to form a complete human being. Education is also one of the determinants in improving human resources. Teachers as educators play an important role in this matter. One of the efforts that can be made to make it happen is by developing learning designs so as to create effective and efficient learning. The Dick and Carry learning design model is one of the learning models that can be used in developing Indonesian language teaching materials at the elementary school level. This study aims to find out how to develop teaching materials for grade 3 Indonesian materials in elementary schools. Based on empirical results and observations that have been made on the Dick and Carry learning model, it is known that the development of this model is very appropriate to be applied in developing Indonesian language teaching materials, especially at the Grade 3 Elementary School level. This is because the Dick and Carry learning model refers to general stages learning development system, so that this model is appropriately applied in skill-based subjects. In addition to this, the Dick and Carry model has 10 systematic learning steps, from identifying the general learning objectives to carrying out evaluations.

Design and implementation of machine learning models to classify and mitigate muras of a micro-LED display

Design and implementation of machine learning models to classify and mitigate muras of a micro-LED display

Food anaphylaxis diagnostic marker compilation in machine learning design and validation.

Traditional food allergy assessment of anaphylaxis remains limited in accuracy and accessibility. Current methods of anaphylaxis risk assessment are costly with low predictive accuracy. The Tolerance Induction Program (TIP) for anaphylactic patients undergoing TIP immunotherapy produced large-scale diagnostic data across biosimilar proteins, which was used to develop a machine learning model for patient-specific and allergen-specific anaphylaxis assessment. In explanation of construct, this work describes the algorithm design for assignment of peanut allergen score as a quantitative measure of anaphylaxis risk. Secondarily, it confirms the accuracy of the machine learning model for a specific cohort of food anaphylactic children. Machine learning model design for allergen score prediction utilized 241 individual allergy assays per patient. Accumulation of data across total IgE subdivision served as the basis of data organization. Two regression based Generalized Linear Models (GLM) were utilized to position allergy assessment on a linear scale. The initial model was further tested with sequential patient data over time. A Bayesian method was then used to improve outcomes by calculating the adaptive weights for the results of the two GLMs of peanut allergy score prediction. A linear combination of both provided the final hybrid machine learning prediction algorithm. Specific analysis of peanut anaphylaxis within one endotype model is estimated to predict the severity of possible anaphylactic reaction to peanut with a recall of 95.2% on a dataset of 530 juvenile patients with various food allergies, including but not limited to peanut allergy. Receiver Operating Characteristic analysis yielded over 99% AUC (area under curve) results within peanut allergy prediction. Machine learning algorithm design established from comprehensive molecular allergy data produces high accuracy and recall in anaphylaxis risk assessment. Subsequent design of additional food protein anaphylaxis algorithms is needed to improve the precision and efficiency of clinical food allergy assessment and immunotherapy treatment.

Model Desain Pembelajaran Pendidikan Agama Islam di SMP Negeri 1 Kadipaten Kabupaten Tasikmalaya

This study aims to find an overview of the PAI learning design model in SMPN 1 Kadipaten Tasikmalaya Regency. This research uses a qualitative approach and the data sources are PAI teachers, class teachers and principals. This study used observation and interview techniques to collect data related to the research objectives. After the data is collected, the next stage is to analyze the data, in this case the researcher uses the approach of Miles et al, namely, data selection and data mapping, data presentation and perification and conclusions. This study found the following conclusions: the learning model applied by PAI teachers is the first active and cooperative learning model, which is a learning model that brings students to be active and creative, not only listening, but also carrying out activities such as processing, finding, and utilizing information. Second, the contextual learning model, this model helps educators relate the material taught to the real-world situation of students, thus allowing learners to find a connection between the knowledge they have and its application in their lives as family and community members.

Media Platform Moodle Assisted Project-Based Learning Model for Short Story Text Learning in Class IX of Junior High School

The purpose of this study was to explain the design of a project-based learning model and the results of its implementation with the help of the Moodle platform in learning short story texts in class IX. The method used is descriptive qualitative method. This study uses a qualitative approach with the ADDIE model development method. Data collection was in the form of teacher and student activities in every aspect of learning model design, short story text assessment instruments, student worksheets and short story text product presentations. The form of data in this study is in the form of scores for short story text assessment sheets. From the results of the research and discussion, it can be concluded that the ability to write short stories in class IX at SMP NU Darul Ma'arif Kaplongan, with a total of 28 students, obtained an average total score of 90.3 out of the highest score of 100. Meanwhile, the field trial phase was carried out at SMPN 1 Karangampel Indramayu which totaling 31 students with an average total score of 80.5 out of the highest score of 100. And SMP Muhammadiyah Karangampel with a total of 31 students obtained an average total score of 80 out of the highest score of 100. Overall implementation results both small group trials and field trials received an average -the average total score is 86.6, it is included in the "Very Eligible" category. That is, the design of a project-based learning model assisted by the Moodle platform media is able to improve students' ability to compose short story texts. Which illustrates an increase in effectiveness and value scores after using the project-based learning model. The design of the Moodle Platform-Assisted Project-Based Learning Model in Learning Short story text is a combination of PjBL and the Moodle Media Platform, so the design of this model is very much needed by teachers in learning Short Story texts. From the implementation of the use of the PjBL model assisted by the Moodle Media Platform, it can be concluded that the developed Project-Based Learning Model Assisted by the Moodle Media Platform in Learning Short Story Texts in SMP/MTS is very feasible for teachers and students to use in learning.

Perencanaan Proses Pembelajaran Taḥsīn dan Taḥfīẓ al-Qur’an di Sekolah Dasar Islam Annajah Bekasi

This study aims to analyze the planning of the learning process of taḥsīn and taḥfīẓ Al-Qur'an in Annajah Bekasi Islamic Elementary School (SD), as well as the constraints in the learning process. The method used in this research is descriptive qualitative type. Primary data sources were taken from several respondents, namely school principals, curriculum departments, and representatives of taḥsīn and taḥfīẓ teachers randomly. Secondary data sources are from the Al-Qur'an, hadith, journals, books, and others. The background that prompted this research is that Annajah Bekasi Islamic Elementary School (SD) is an elementary level education institution that has serious concern for learning the Qur'an with the Nurul Bayan Method and the Asy-Syafi'i Method. The results of this study are: (1) the planning of the learning process for taḥsīn and taḥfīẓ Al-Qur'an in Annajah Bekasi Islamic Elementary School (SD) is generally neat and orderly according to the learning design model of Dick and Carrey (1985 ), which consists of 10 (ten) systematic learning steps, starting from identifying general learning objectives to designing and carrying out summative evaluations, and (2) some obstacles in the learning process of taḥsīn and taḥfīẓ Al-Qur'an in Annajah Bekasi Islamic Elementary School (SD), among others, that there are students who stammer in reading the Qur'an, lack mastery of the rules of recitation, do not recognize hijā'iyyah letters, and lack motivation in learning. The difference with previous studies is that the focus of the discussion in this research is more broadly related to the analysis of planning the learning process of taḥsīn and taḥfīẓ Al-Qur'an, as well as obstacles in the learning process at Annajah Bekasi Islamic Elementary School (SD) in under the auspices of the Ministry of Education and Culture by using the learning design model Dick and Carrey (1985) as an analytical knife, and the method used refers to the Nurul Bayan Method and the Asy-Syafi'i Method.

Project Based Learning Model Digital Media Assistance in Learning Persuasive Speech Text in SMP/MTs

The purpose of this research is to explain the design of project-based learning model with the help of digital media in teaching persuasive speech texts in SMP/MTs. The method used is descriptive qualitative method. This study uses a qualitative approach with the ADDIE model development method. Data collection was in the form of teacher and student activities in every aspect of learning model design, speech text assessment instrument, viewing instrument sheets and presentation of speech text products. The form of data in this study is in the form of pretest and posttest scores for speech text assessment sheets. From the results of the research and discussion, it can be concluded that the ability to write speech texts in the pretest and posttest obtained an average score of 68,70 in pretest and 86,24 in the posttest, including in the good category. The result of the analysis of the N-gain score are 0,5635 in the range 0.3 ≤ g ≤ 0.7 in the medium category. From user/educator data on the implementation of the digital media-assisted learning model design used with interpretation criteria an average of 89% is included in the very feasible category. That is, the design of a project-based learning model assisted by digital media is able to improve students’ abilities in making speech texts. Which illustrates an increase in effectiveness and value scores after using the project-based learning model. The design of the Digital Media Assisted Project-Based Learning Model in Persuasive Speech Text Learning is a combination of Project Based Learning and Digital media, so the design of this model is very much needed by teachers in persuasive Speech text learning. From the implementation of the use of the digital media-assisted Project based Learning model, it can be concluded that the developed digital media-assisted project-based learning model in learning persuasive speech texts in SMP/MTs is very feasible for teachers and students to use in learning.

TPACK-based hybrid learning model design for computational thinking skills achievement in mathematics

Hybrid learning implementation is closely related with technology. In designing hybrid lessons, lecturers need to have both the pedagogical and content-related skills that make the best use of technology so that it also improves the students’ skills, including their computational thinking skills. The purpose of this research is to examine whether a hybrid learning model based on Technology Pedagogy and Content Knowledge (TPACK) affects the computational thinking skills achievement in a valid, practical, and effective manner. This research is development research using the Plomp’s model and is described in a qualitative descriptive manner. The participants were lecturers and 38 first-year students at Primary School Teacher Education Study Program of Padang State University. The instruments were validation sheets, practicality assessment sheets by lecturers, questionnaire sheets for students, and computational thinking skills test questions. The results of the validity test show that in terms of content, language, and e-learning practice, they all met very valid criteria with an average percentage score of 85.9%, 86.2%, and 84.1%, respectively. The practicality test results of model handbooks, Semester Lesson Plans, Lecture Program Units, computational thinking questions, additional material, and e-learning model meet the practical criteria with an average percentage score of 77.5%, 86.8%, 89.1%, 83.8%, 79%, and 82.4%, respectively. The average percentage of student assessment scores is 82.9% and meets the practical criteria. The results of the effectiveness test showed that 21.9% of students had reached the moderate level, 6.3% reached the low level, and 71.9% reached the very low level. This study recommends future researchers to develop a hybrid learning model based on TPACK to achieve other 21st-century abilities.

Leveraging conscious and nonconscious learning for efficient AI.

Various interpretations of the literature detailing the neural basis of learning have in part led to disagreements concerning how consciousness arises. Further, artificial learning model design has suffered in replicating intelligence as it occurs in the human brain. Here, we present a novel learning model, which we term the "Recommendation Architecture (RA) Model" from prior theoretical works proposed by Coward, using a dual-learning approach featuring both consequence feedback and non-consequence feedback. The RA model is tested on a categorical learning task where no two inputs are the same throughout training and/or testing. We compare this to three consequence feedback only models based on backpropagation and reinforcement learning. Results indicate that the RA model learns novelty more efficiently and can accurately return to prior learning after new learning with less computational resources expenditure. The final results of the study show that consequence feedback as interpretation, not creation, of cortical activity creates a learning style more similar to human learning in terms of resource efficiency. Stable information meanings underlie conscious experiences. The work provided here attempts to link the neural basis of nonconscious and conscious learning while providing early results for a learning protocol more similar to human brains than is currently available.

Experience Based Learning Model in Developing Creativity Through Poetry Text Material in SMA/SMK

This study aims to develop an experience-based learning model in developing creativity through poetry text material in SMA/SMK. Experience-based learning models are often used by teachers to develop language skills through an appropriate material. Creativity is an ability that students need to navigate life in the millennial era. This ability needs to be combined with an experience-based learning model so that students' creativity in poetry is growing. This research uses a research and development method with the ADDIE model (Analysis Design Development Implementation Evaluate). The steps of this research consist of, needs analysis, designing learning models, developing learning models, product implementation. The method of collecting data in this study was by collecting data from the results of expert validation questionnaires and student assessment questionnaires. Data acquisition was analyzed descriptively by using a Likert scale. The results of the design of experience-based learning models in the development of creativity through poetry text material. The implementation of experiential learning models in the development of creativity through poetry text material in SMA/SMK as a whole received a positive response in the "Very Eligible" category for use as teaching material for Indonesian language subjects in poetry texts.

Design of Flipped Classroom Learning in Vocational Education to improve HOT Skills in the era of the Industrial Revolution 4.0

This study aims to describe and analyze how the flipped classroom learning model design in Computer Network Security learning in Information Systems Study Program, Faculty of Computer Science. This study aims to design and develop a flipped classroom model to improve 21st century learning ability (4C).The type of development method used is Research and Development (R & D). The development procedure used is 4-D, and is part of the development stage of a learning model. The type of data used is primary data where data is obtained directly from universities, lecturers, experts and students. The results of research based on needs analysis and literature reviews have produced a flipped classroom model design, and based on the results of expert validation it can be concluded that the website-based flipped classroom design is considered quality to measure student competency achievement consisting of menu material, student recap, questions, concept understanding, assessment self, character values, and recap values. Based on the validation results of all the design indicators for the flipped classroom model which were assessed by the validator from the aspects of the feasibility of the content, the language of the instruments, and the feasibility of the graphic aspects of the valid and feasible categories for testing.
 Keywords: HOT skills; flipped; classroom learning; vocational education

Deep learning model of convolutional neural networks powered by a genetic algorithm for prevention of traffic accidents severity

The World Health Organization highlights that the number of annual road traffic deaths has reached 1.35 million (Global Status Report on Road Safety 2018). In addition, million of people suffer more or less important injuries as a consequence of this type of accidents. In this scenario, the prediction of the severity of traffic accidents is an essential point when it comes to improving the prevention and reaction of the entities responsible. On the other hand, the development of reliable methodologies to predict and classify the level of severity of traffic accidents, based on various variables, is a key component in the field of research in road safety. This work aims to propose a new approach, based on convolutional neural networks, for the detection of the severity of traffic accidents. Behind this objective is the preprocessing, analysis and visualization of data as well as the design, implementation and comparison of machine learning models considering accuracy as a performance indicator. For this purpose, a scalable and easily reusable methodology has been implemented. This methodology has been compared with other deep learning models verifying that the results of the designed neural network offer better performance in terms of quality measures.

Toward improved prediction of recycled brick aggregate concrete compressive strength by designing ensemble machine learning models

The utilization of recycled brick aggregate concrete (RBC) is an area of active research, in which further investigation is needed to develop accurate models for predicting the behavior of RBC to ensure its safe and sustainable use in construction. This study presents an effective way to determine RBC's compressive strength (CS) based on an appropriate design of ensemble machine learning (ML) models, namely Gradient Boosting, Light Gradient boosting, AdaBoost, Extreme Gradient Boosting, Stacking, and Voting. A database covering 393 test results is compiled from the relevant literature for training and testing the models. In addition, 10-fold cross-validation and random splitting of data are used to ensure the reliability of predictions, as well as prevent overfitting. The findings reveal that the Stacking model has the greatest predictive ability, with a coefficient of determination of 0.95, root mean square error of 2.74 MPa, mean absolute error of 2.09 MPa, mean absolute percentage error of 0.10 on the testing dataset. In addition, Feature importance and Partial dependence plots analysis are utilized to investigate the impact of each RBC component on its CS. In RBC’s mixed-component design, the outcomes of this research, along with a developed Graphical User Interface (GUI) for the CS, might be of great use to material engineers.