Design Of Learning Model Research Articles

Design and Validation of a Deep Learning Model for Renal Stone Detection and Segmentation on Kidney-Ureter-Bladder Images.

Kidney-ureter-bladder (KUB) imaging is used as a frontline investigation for patients with suspected renal stones. In this study, we designed a computer-aided diagnostic system for KUB imaging to assist clinicians in accurately diagnosing urinary tract stones. The image dataset used for training and testing the model comprised 485 images provided by Kaohsiung Chang Gung Memorial Hospital. The proposed system was divided into two subsystems, 1 and 2. Subsystem 1 used Inception-ResNetV2 to train a deep learning model on preprocessed KUB images to verify the improvement in diagnostic accuracy with image preprocessing. Subsystem 2 trained an image segmentation model using the ResNet hybrid, U-net, to accurately identify the contours of renal stones. The performance was evaluated using a confusion matrix for the classification model. We conclude that the model can assist clinicians in accurately diagnosing renal stones via KUB imaging. Therefore, the proposed system can assist doctors in diagnosis, reduce patients' waiting time for CT scans, and minimize the radiation dose absorbed by the body.

The Effectiveness of Educational Tools Grounded in Local Knowledge in Enhancing High School Students' Creative Thinking Skills in the Context of Sound Waves

The objective of this study was to describe the effectiveness of learning tools based on local wisdom in enhancing creative thinking skills in the context of sound waves. The research was conducted at SMA N 1 North Gorontalo, specifically targeting eleventh-grade students. The study employed a Research and Development (R&D) approach, utilizing the ADDIE learning design model, which encompasses the stages of Analysis, Design, Development, Implementation, and Evaluation. However, it is important to note that this study focuses solely on describing the effectiveness of the learning materials, without delving into other aspects. The effectiveness of the learning tools was assessed during the implementation stage. The findings of the study indicated that student engagement during the two sessions was rated as very good. The assessment of attitudes in both sessions resulted in a percentage of 87.89% and 89.85% respectively, falling within the very good category. Similarly, the assessment of skills exhibited percentages of 85.70% and 89.85% respectively, also within the very good range. Furthermore, the test conducted to evaluate students' creative thinking abilities revealed that the majority of students demonstrated creativity in utilizing traditional polopalo musical instruments. These findings suggest that learning tools based on local wisdom have the potential to enhance students' creative thinking skills.

Design of a high-density bio-inspired feature analysis deep learning model for sub-classification of natural & synthetic imagery

Design of a high-density bio-inspired feature analysis deep learning model for sub-classification of natural & synthetic imagery

Development of a Problem-Based Learning Model Integrated with 21st Century Skills

This research aims to develop the PBL-4C Learning Model to promote mathematical problem-solving skills among students and to create a valid, practical, and effective PBL-4C learning model. It falls under the category of research and development (R&D). The model development process follows the Plomp development model, which comprises four phases: (1) initial investigation phase, involving student needs analysis, task analysis, concept analysis, and learning objective formulation; (2) design phase, encompassing the design of the PBL-4C learning model, including model syntax, social principles, reaction principles, instructional impact, facilitation, and support system; (3) implementation phase, which involves the complete construction of the PBL-4C model based on the previous design phase; and (4) testing, evaluation, and revision phase, which includes validation and pilot testing. The validation results of various validity instruments include the average score of the PBL-4C learning model book (3.67), student’s book (3.98), student’s worksheets (4.35), lesson plans (4.48), learning outcome tests (3.71), students’ response questionnaires on the implementation of the PBL-4C learning model (4.12), students’ response questionnaires on the student’s worksheets (3.73), students’ response questionnaires on the student’s book (3.75), teacher’s response questionnaires (3.65), students’ activity observation sheets (3.75), and learning implementation observation sheets (3.40). Each of these validity instruments falls under the highly valid category, indicating that the PBL-4C learning model book and its supporting instruments meet the validity criteria. The PBL-4C model product is effectively implemented in teaching mathematics, specifically the topic of systems of linear equations with two variables. The effectiveness of the PBL-4C learning model is evaluated based on data analysis of its effectiveness components: (a) the average percentage of student activities during the implementation of the PBL-4C model in pilot test I is 63.25%, which falls under the "good" category, and the average percentage of student activities in pilot test II is 79%, falling under the "good" category; (b) the average percentage of student responses in pilot test I is 76% with positive responses, and in pilot test II, it is 82% with positive responses; (c) the learning outcome test results indicate that in pilot test I, 80% of students achieved mastery, while in pilot test II, 92% of students achieved mastery, thus meeting the criteria for classical mastery.

Design of Learning Models and Algorithmic Approaches in an Expert Learning System

Design of Learning Models and Algorithmic Approaches in an Expert Learning System

소통강화를 위한 교양 스페인어 문제중심학습 모형

This study aimed to design a problem-based learning model for general Spanish course to strengthen communication using learner-centered teaching methods. The subjects of this study were divided into two classes for a total of 80 freshmen students taking Spanish classes in a university in Daejeon City for 10 weeks in the second semester of 2022. After reviewing the validity of five experts, the final model was completed through learner evaluation after applying it to actual classes. The results of the expert validity test showed that the average of suitability, explanatory power, and usefulness items was 3.0 or higher, but the average of comprehension items was 2.8 and less than 3.0. The final model of problem-based learning design was completed by reflecting the results of expert validity verification and learner evaluation.

MBFRDH: Design of a Multimodal Bioinspired Feature Representation Deep Learning Model for Identification of Heart-Diseases

Electrocardiograms (ECGs) are generated by checking different beating patterns of heart, and are widely used for identification of multiple heart-related issues. Existing deep learning models that are proposed for ECG analysis are either highly complex, or showcase lower scalability when applied to clinical scans. To overcome these issues, this text proposes design of a novel multimodal bioinspired feature representation deep learning model for identification of heart-diseases. The proposed model initially collects large-scale ECG datasets, and extracts Fourier, Cosine, iVector, Gabor, and Wavelet components. These components are given to a Grey Wolf Optimization (GWO) based feature selection model, which assists in identification of high-inter-class variance feature sets. This is done via modelling a variance-based fitness function and fusing it with an Iterative Learning Model (ILM) that use feedback-accuracy levels for optimization of selected feature sets. The extracted features are used to incrementally train a custom 1D Binary-Augmented Convolutional Neural Network (1D BACNN) that can be trained for multiclass scenarios. The BACNN Model is trained individually for each of the heart diseases. Each BACNN categorizes input ECG samples between ‘Normal’, and ‘Heart-Disease’ categories. Due to use of this binary-type classification, the proposed model is able to achieve a consistent 99.9% accuracy for multiple heart disease sets, which is found to be higher than most of the existing multiclass techniques. The model was tested for Angina, Arrhythmia, Valve disease, and Congenital heart conditions, and was observed to achieve 3.5% higher precision, 4.9% higher accuracy, with 1.2% increase is computational delay, which makes it highly suitable for real-time clinical use cases.

Backward Design : Strategi Pembelajaran dalam Meningkatkan Keterlibatan Siswa

This research aims to explore the backward design model as a learning strategy to enhance student engagement. This research adopts a descriptive approach with a qualitative method. The participants of this study were high school teachers selected through the snowball sampling technique. The data collection was conducted through interviews and participatory observation, while data analysis follows an interactive model analysis approach, including data reduction, data display, and conclusion drawing/verification. The research findings indicated that the backward design was one of the effective learning design models as it focused on understanding and developing students' thinking skills. The model directed teachers to set learning objectives that wereappropriate to the student's needs and skills, conduct assessments to measure students' understanding and skills and formulate when delivering materials or certain concepts gradually. Implementing concept-checking techniques allows teachers to measure students' understanding periodically, and applying interaction patterns accommodates continuous interactions.

Development of Dance Learning Model at School

This research was motivated because the learning activities of cultural arts at SMA Negeri 1 Koto Balingka were not carried out correctly. The teaching and learning process was carried out less than optimally, extracurricular activities did not go well. Therefore, this study aimed to 1) find out the model used in class IPA SMA N 1 Koto Balingka, 2) produce a dance learning model design in school, 3) obtain a dance learning model in school and 4) produce a dance learning model in school. This research was development research using a 4D development model (define, design, develop, and disseminate). The type of data generated was a product trial assessing the creative movement resulting from the educational dance learning model. The instrument used to collect data consisted of a validity questionnaire by validators/experts/experts. The data analysis techniques in this study were product validity analysis, product practicality analysis, product effectiveness analysis. The research results could be concluded that learning to design dance works using the Development of Educational Dance Learning Model based on Themes (RIDAMA) is very attractive to students. Dance learning activities were easy to understand. It could be seen from the value of the model's validity of 90.26 with very valid criteria. The value of the practicality of the model by the teacher was 93.4 with very practical criteria. In comparison, students' practicality of the model was 98 very practical criteria. The model's effectiveness consisted of an overall assessment of student activity with a score of 92 with very active criteria and student learning outcomes in the knowledge aspect with a score of 97 with a very good predicate. The study results concluded that the Theme-based Educational Dance Learning Model (MPTPbT) developed was valid, practical, and effective. This model can overcome the limitations of dance learning tools in schools.

Development of Dance Learning Model at School

This research was motivated because the learning activities of cultural arts at SMA Negeri 1 Koto Balingka were not carried out correctly. The teaching and learning process was carried out less than optimally, extracurricular activities did not go well. Therefore, this study aimed to 1) find out the model used in class IPA SMA N 1 Koto Balingka, 2) produce a dance learning model design in school, 3) obtain a dance learning model in school and 4) produce a dance learning model in school. This research was development research using a 4D development model (define, design, develop, and disseminate). The type of data generated was a product trial assessing the creative movement resulting from the educational dance learning model. The instrument used to collect data consisted of a validity questionnaire by validators/experts/experts. The data analysis techniques in this study were product validity analysis, product practicality analysis, product effectiveness analysis. The research results could be concluded that learning to design dance works using the Development of Educational Dance Learning Model based on Themes (RIDAMA) is very attractive to students. Dance learning activities were easy to understand. It could be seen from the value of the model's validity of 90.26 with very valid criteria. The value of the practicality of the model by the teacher was 93.4 with very practical criteria. In comparison, students' practicality of the model was 98 very practical criteria. The model's effectiveness consisted of an overall assessment of student activity with a score of 92 with very active criteria and student learning outcomes in the knowledge aspect with a score of 97 with a very good predicate. The study results concluded that the Theme-based Educational Dance Learning Model (MPTPbT) developed was valid, practical, and effective. This model can overcome the limitations of dance learning tools in schools.

The Design and Optimization of an Acoustic and Ambient Sensing AIoT Platform for Agricultural Applications.

The use of technology in agriculture has been gaining significant attention recently. By employing advanced tools and automation and leveraging the latest advancements in the Internet of Things (IoT) and artificial intelligence (AI), the agricultural sector is witnessing improvements in its crop yields and overall efficiency. This paper presents the design and performance analysis of a machine learning (ML) model for agricultural applications involving acoustic sensing. This model is integrated into an efficient Artificial Intelligence of Things (AIoT) platform tailored for agriculture. The model is then used in the design of a communication network architecture and for determining the distribution of the computing load between edge devices and the cloud. The study focuses on the design, analysis, and optimization of AI deployment for reliable classification models in agricultural applications. Both the architectural level and hardware implementation are taken into consideration when designing the radio module and computing unit. Additionally, the study encompasses the design and performance analysis of the hardware used to implement the sensor node specifically developed for sound classification in agricultural applications. The novelty of this work lies in the optimization of the integrated sensor node, which combines the proposed ML model and wireless network, resulting in an agricultural-specific AIoT platform. This co-design enables significant improvements in the performance and efficiency for acoustic and ambient sensing applications.

DeeP4med: deep learning for P4 medicine to predict normal and cancer transcriptome in multiple human tissues

BackgroundP4 medicine (predict, prevent, personalize, and participate) is a new approach to diagnosing and predicting diseases on a patient-by-patient basis. For the prevention and treatment of diseases, prediction plays a fundamental role. One of the intelligent strategies is the design of deep learning models that can predict the state of the disease using gene expression data.ResultsWe create an autoencoder deep learning model called DeeP4med, including a Classifier and a Transferor that predicts cancer's gene expression (mRNA) matrix from its matched normal sample and vice versa. The range of the F1 score of the model, depending on tissue type in the Classifier, is from 0.935 to 0.999 and in Transferor from 0.944 to 0.999. The accuracy of DeeP4med for tissue and disease classification was 0.986 and 0.992, respectively, which performed better compared to seven classic machine learning models (Support Vector Classifier, Logistic Regression, Linear Discriminant Analysis, Naive Bayes, Decision Tree, Random Forest, K Nearest Neighbors).ConclusionsBased on the idea of DeeP4med, by having the gene expression matrix of a normal tissue, we can predict its tumor gene expression matrix and, in this way, find effective genes in transforming a normal tissue into a tumor tissue. Results of Differentially Expressed Genes (DEGs) and enrichment analysis on the predicted matrices for 13 types of cancer showed a good correlation with the literature and biological databases. This led that by using the gene expression matrix, to train the model with features of each person in a normal and cancer state, this model could predict diagnosis based on gene expression data from healthy tissue and be used to identify possible therapeutic interventions for those patients.

Neural architecture search for resource constrained hardware devices: A survey

AbstractWith the emergence of powerful and low‐energy Internet of Things devices, deep learning computing is increasingly applied to resource‐constrained edge devices. However, the mismatch between hardware devices with low computing capacity and the increasing complexity of Deep Neural Network models, as well as the growing real‐time requirements, bring challenges to the design and deployment of deep learning models. For example, autonomous driving technologies rely on real‐time object detection of the environment, which cannot tolerate the extra latency of sending data to the cloud, processing and then sending the results back to edge devices. Many studies aim to find innovative ways to reduce the size of deep learning models, the number of Floating‐point Operations per Second, and the time overhead of inference. Neural Architecture Search (NAS) makes it possible to automatically generate efficient neural network models. The authors summarise the existing NAS methods on resource‐constrained devices and categorise them according to single‐objective or multi‐objective optimisation. We review the search space, the search algorithm and the constraints of NAS on hardware devices. We also explore the challenges and open problems of hardware NAS.

Machine learning in predicting anti-money laundering compliance with protection motivation theory among professional accountants

Money laundering represents a significant global threat, necessitating the vigilance of professional accountants in detecting and reporting suspicious customer activities within their jurisdiction to the relevant authorities. Despite the legal obligation to comply with anti-money laundering regulations, professional accountants' adherence to these measures remains insufficient. Previous research on machine learning techniques for combating money laundering has predominantly concentrated on predicting suspicious transactions, rather than evaluating compliance behavior. This study aims to develop a machine learning prediction model to assess the inclination of professional accountants towards adhering to anti-money laundering regulations, serving as an early signal system to gauge their willingness to abide by the law in their professional responsibilities. The research elaborates on the design and implementation of machine learning models based on three algorithms: Decision Tree, Gradient Boosted Tree, and Support Vector Machine. The paper offers two types of comparisons from distinct perspectives: firstly, the performance of each algorithm in predicting real cases of anti-money laundering compliance, and secondly, the contribution of attributes measured by weights of correlation in different algorithms. Alongside demographic factors, the study evaluates the effectiveness of each algorithm in anti-money laundering compliance by utilizing five attributes derived from the Protection Motivation Theory (PMT). The findings demonstrate the significance of all attributes, including demography and PMT, in all machine learning models, with both Gradient Boosted Tree and Support Vector Machine achieving a proportion of variance of 0.8 or higher. This indicates the potential of these algorithms in effectively measuring and predicting professional accountants' intentions to comply with anti-money laundering regulations.

Development of Natural Material-Based Art-Based Learning Model to Stimulate Aesthetic Experiences in Early Children

The purpose of this study is to analyze needs, develop designs, and analyze the effectiveness of nature-based art-based learning models for stimulating aesthetic experiences in early childhood in Kindergarten, Banjarsari District, Surakarta City. This research is a type of R&D research and development. The results of the study: 1) teachers and students need a learning model based on natural materials with an art approach so that they can stimulate children's aesthetic experiences. Learning models that can be implemented easily and do not require a lot of money by utilizing nature and the surrounding environment; 2) the design of art-based learning models based on natural materials includes learning natural-based painting, learning natural-based collage art, learning natural-based handicraft art, learning natural-based batik art, learning natural-based color art; 3) the art-based learning model based on natural materials is effective in stimulating children's aesthetic experience. In this case, the value of developing an art-based learning model based on natural materials is effective in stimulating children's aesthetic experience.

The Effect of the Empirical Base Nature of Science Learning Model on the Understanding of Nature of Science in Empirical Aspect

Understanding of the nature of science, especially the empirical aspect, is now receiving less attention and seriousness from the educational community. The aim of this study was determine the effect of the Empirical Base Nature of Science (EBNOS) learning model on understanding NOS, especially on the empirical aspects of elementary school students. This study used a pre-experimental method through a one group pretest posttest design. The research subjects selected by convenience sampling were 51 students of class V of a private elementary school in the city of Bandung. The instrument used to collect data related to students' understanding of NOS at the pretest and posttest uses a Likert scale. Additional aspects explored from research subjects to support research results are aspects of gender and interest in science. Quantitative descriptive method with the help of SPSS and Microsoft Excel was used to analyze the data collected from the questionnaire. The results showed that the application of the Empirical Base Nature of Science (EBNOS) learning model design had a significant effect on increasing students' understanding of NoS, especially on the empirical aspect. However, understanding of the empirical aspect of NOS was not influenced by the gender of the students or their interest in science itself. This research is expected to contribute to the improvement of the learning process which aims to achieve science learning goals on the NOS dimension, especially on the empirical aspect. Suggestions for further research are developing specific learning models that can improve understanding of NOS in other aspects.

Self-Regulated Learning In Mathematics Learning: Design of Learning Model Based on Bibliometric Analysis

Self-Regulated Learning (SRL) has been a growing topic since 1983 in the Scopus database. The importance of SRL has been known through several studies because it is proven to be able to improve student learning outcomes, both at school and in life. The purpose of this study is to determine the trend of topics in the field, through bibliometric analysis with the help of the VOSViewer application. Findings reveal that there is still little discussion on the topic of SLR in mathematics learning. The journal that can be used as a reference is the International Journal of Science and Mathematics Education, which the latest topics related to Learning Systems and Education. Suggestions for further research are to make technology-based learning models using the latest theories and research results and evaluate the results achieved for improving students’ learning outcomes in mathematics and in life.

DEVELOPMENT OF A STORY LEARNING MODEL USING HAND POOLS IN ELEMENTARY SCHOOLS

Abstract
 This study aims to analyze the need to use learning and design learning models and test the validity of storytelling learning models. The method used in this research is R&D (Research and Development) with the ADDIE development model. This research was conducted in class IV SDN 60 Salubattang, Palopo City. Data collection techniques using interviews, questionnaires, and observation. Data analysis techniques were performed by classifying the data obtained into qualitative and quantitative data. The needs analysis results illustrate that students need interesting concrete visualizations to increase interest in reading. The design of a storytelling learning model assisted by hand puppet media is described in a guidebook which contains model components, media and how to make them, and the technical application of the learning model. The learning model developed through the validation stage by design experts, material experts, linguists, and elementary school instructional practice experts obtained results that were feasible to use.

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.