Active Learning Techniques Research Articles

Deep learning based active learning technique for data annotation and improve the overall performance of classification models

Deep learning (DL) models achieved a lot of success due to the availability of labeled training data. In contrast, labeling a huge amount of data by a human is a time-consuming and expensive solution. Active Learning (AL) efficiently addresses the issue of labeled data collection at a low cost by picking the most useful samples from a large number of unlabeled datasets. However, current AL techniques largely depend on regular human involvement to annotate the most uncertain/informative samples in the collection. Therefore, we introduce a Deep Learning Based Active Learning (DLBAL) method that may incrementally learn from a small number of annotated training samples to build an effective classifier with optimum feature representation. The proposed DLBAL method improves the contemporary AL approaches in two steps. First, EfficientNet-B0 networks are integrated into AL. The incrementally annotated useful samples may be used to update both the representation of the feature and the classification algorithm concurrently. Second, we provide an effective technique for sample selection that improves the performance of the classifier with fewer manual annotations at a lower cost. In contrast to previous approaches that emphasize only uncertain instances with low prediction confidence, we obtain a significant number of instances with high-confidence from the unannotated set for pattern learning. Particularly, these instances are randomly picked and assigned labels to them continuously. Detailed experiments are conducted to show that the proposed DLBAL system obtained reasonable results on two difficult image classification datasets, namely Cross-Age Celebrity face recognition Dataset (CACD) and Caltech-256.

Innovative Approaches for Preparing Special Education Preservice Teachers

For years, there has been a shortage of educators qualified to teach students with disabilities. The effect this has on student outcomes is immeasurable. To overcome this shortage, universities are searching for ways to enroll more students into their special education programs, ensure these graduates are prepared for long-term employment, and arm them with the skills necessary to best prepare their future learners. To this end, special education teacher educators seek ways to instruct most effectively during their limited time with their preservice teachers. This article discusses an instructional method to help teacher educators accomplish this goal, drawing on theoretical frameworks related to active learning techniques. Compared to the traditional method of instruction, the blended learning approach affords teachers more in-class time to actively engage preservice teachers with their course content while maintaining a rigorous learning environment. The authors explain how this model can be incorporated into synchronous and asynchronous courses and share valuable online educational resources for successful implementation. Additionally, the authors discuss active learning strategies and video analysis tools to support preservice teachers in both the classroom and during field supervision. All of these focus on equipping preservice teachers to effectively handle the diverse and constantly evolving demands of the contemporary classroom, which may positively impact teacher retention and create a more stable teaching workforce.

Rating sovereign credit risk: A simulation for advanced economics and finance students

The authors of this article propose a classroom simulation designed for advanced economics or finance courses whereby student teams role-play Moody’s sovereign credit risk analysts. Despite the importance of sovereign credit risk ratings in affecting the funding liquidity of countries, the process generating ratings is a black box. The authors use active and experiential learning techniques to guide student teams in mimicking the process used by Moody’s analysts to assign a sovereign credit risk rating to one of 12 diverse countries. An accompanying YouTube video guides students in navigating three Web sites to retrieve macroeconomic data informing sovereign credit risk ratings. The simulation may be utilized in face-to-face and synchronous online environments.

EFL TEACHERS’ PERCEPTIONS ABOUT IMPLEMENTING ACTIVE LEARNING TECHNIQUES IN TEACHING SPEAKING

Active learning has indicated its important role in learners’ speaking process and involved students in classroom activities. Therefore, there is a pressing need for implementing active learning techniques to engage students in speaking English. The research entitled “The EFL Teachers’ Perceptions about Implementing Active Learning Techniques in Teaching Speaking” was conducted to obtain two research aims, including (1) to find out teachers’ perceptions about implementing active learning techniques in their speaking classrooms; (2) to investigate the benefits and difficulties of implementing active learning techniques in teaching speaking. This is descriptive research using mix-methods. The participants were 56 teachers who are from two English language centers in Can Tho City responding to the questionnaire. Besides, semi-structured interviews were also administrated to investigate the benefits and difficulties teachers faced when implementing active learning techniques in teaching speaking. The findings show that teachers had positive perceptions of implementing active learning techniques in their speaking classrooms in the Mekong Delta and provided some valuable information for further topic-related research.<p> </p><p><strong> Article visualizations:</strong></p><p><img src="/-counters-/soc/0024/a.php" alt="Hit counter" /></p>

A mixed method analysis of student satisfaction with active learning techniques in an online graduate anatomy course: Consideration of demographics and previous course enrollment.

Online learning has become an essential part of mainstream higher education. With increasing enrollments in online anatomy courses, a better understanding of effective teaching techniques for the online learning environment is critical. Active learning has previously shown many benefits in face-to-face anatomy courses, including increases in student satisfaction. Currently, no research has measured student satisfaction with active learning techniques implemented in an online graduate anatomy course. This study compares student satisfaction across four different active learning techniques (jigsaw, team-learning module, concept mapping, and question constructing), with consideration of demographics and previous enrollment in anatomy and/or online courses. Survey questions consisted of Likert-style, multiple-choice, ranking, and open-ended questions that asked students to indicate their level of satisfaction with the active learning techniques. One hundred seventy Medical Science master's students completed the online anatomy course and all seven surveys. Results showed that students were significantly more satisfied with question constructing and jigsaw than with concept mapping and team-learning module. Additionally, historically excluded groups (underrepresented racial minorities) were generally more satisfied with active learning than non-minority groups. Age, gender, and previous experience with anatomy did not influence the level of satisfaction. However, students with a higher-grade point average (GPA), those with only a bachelor's degree, and those with no previous online course experience were more satisfied with active learning than students who had a lower GPA, those holding a graduate/professional degree, and those with previous online course experience. Cumulatively, these findings support the beneficial use of active learning in online anatomy courses.

UAV Hyperspectral Characterization of Vegetation Using Entropy-Based Active Sampling for Partial Least Square Regression Models

Optimization of agricultural practices is key for facing the challenges of modern agri-food systems, which are expected to satisfy a growing demand of food production in a landscape characterized by a reduction in cultivable lands and an increasing awareness of sustainability issues. In this work, an operational methodology for characterization of vegetation biomass and nitrogen content based on close-range hyperspectral remote sensing is introduced. It is based on an unsupervised active learning technique suitable for the calibration of a partial least square regression. The proposed technique relies on an innovative usage of Shannon’s entropy and allows for the set-up of an incremental monitoring framework from scratch aiming at minimizing field sampling activities. Experimental results concerning the estimation of grassland biomass and nitrogen content returned RMSE values of 2.05 t/ha and 4.68 kg/ha, respectively. They are comparable with the literature, mostly relying on supervised frameworks and confirmed the suitability of the proposed methodology with operational environments.

Accelerated discovery of high-performance Al-Si-Mg-Sc casting alloys by integrating active learning with high-throughput CALPHAD calculations

ABSTRACT Scandium is the best alloying element to improve the mechanical properties of industrial Al-Si-Mg casting alloys. Most literature reports devote to exploring/designing optimal Sc additions in different commercial Al-Si-Mg casting alloys with well-defined compositions. However, no attempt to optimize the contents of Si, Mg, and Sc has been made due to the great challenge of simultaneous screening in high-dimensional composition space with limited experimental data. In this paper, a novel alloy design strategy was proposed and successfully applied to accelerate the discovery of hypoeutectic Al-Si-Mg-Sc casting alloys over high-dimensional composition space. Firstly, high-throughput CALculation of PHAse Diagrams (CALPHAD) solidification simulations of ocean of hypoeutectic Al-Si-Mg-Sc casting alloys over a wide composition range were performed to establish the quantitative relation ‘composition-process-microstructure’. Secondly, the relation ‘microstructure-mechanical properties’ of Al-Si-Mg-Sc hypoeutectic casting alloys was acquired using the active learning technique supported by key experiments designed by CALPHAD and Bayesian optimization samplings. After a benchmark in A356-xSc alloys, such a strategy was utilized to design the high-performance hypoeutectic Al-xSi-yMg alloys with optimal Sc additions that were later experimentally validated. Finally, the present strategy was successfully extended to screen the optimal contents of Si, Mg, and Sc over high-dimensional hypoeutectic Al-xSi-yMg-zSc composition space. It is anticipated that the proposed strategy integrating active learning with high-throughput CALPHAD simulations and key experiments should be generally applicable to the efficient design of high-performance multi-component materials over high-dimensional composition space.

Applying active learning method to improve teaching outcomes in architectural engineering courses

PurposeThis study analyzes the effect of the techniques of active teaching and learning as a way of delivery on the outcomes of quality learning. Focusing on the courses of architectural science taught in a nontraditional method using various active learning strategies, the study takes the case study of the course Building Illumination and Acoustics (BIA) delivered in the academic year 2019–2020 at the University of Sharjah (UoS)'s Architectural Engineering Department (AED).Design/methodology/approachUtilizing both quantitative and qualitative research approaches, the study applied a case study and survey as methods. A questionnaire was designed and performed to assess the level of students' satisfaction with the implemented active teaching method.FindingsThe vibrant learning setting made the students actively engaged and more motivated and enthusiastic. The active learning practices used, including employing senses as in sight and hearing, reasoning rationally and intuitively, reflecting and acting, working steadily and in fits and starts, creating mathematical models, visualizing and memorizing and drawing analogies, were efficient in boosting their ability to comprehend theoretical concepts more effectively. The delivery style effectively enhances quality learning when various active techniques are used pedagogically beyond being merely a utilitarian instrument to prepare novice students of architectural engineering to fulfill practical challenges.Research limitations/implicationsThis article focuses specifically on a theoretical, scientific non-studio course in a particular program of architectural engineering in a particular semester before the dramatic changes in styles of teaching delivery that happened due to the COVID-19 pandemic. Future research could further highlight its results by comparing them to statistical evidence of the development of the course, especially for the duration of online teaching during the pandemic and the hybrid teaching period after it.Originality/valueThis article contributes to the development of teaching and learning of architectural engineering in the local Emirati context by putting original theories of teaching into practice. This paper further contributes to the field of architectural pedagogy in terms of the effect of active learning in the architecture field in the non-studio courses in higher education in the United Arab Emirates.

IMPROVING SCIENCE PROCESS SKILLS OF STUDENTS WITH MILD INTELLECTUAL DISABILITIES

Relevant literature has an unexplored question on how prediction-observation-explanation (POE) worksheets affect science process skills of students with mild intellectual disabilities (SMID). Therefore, this research aimed to examine the effect of POE worksheets developed for the “matter and its nature” subject on SMID’s science process skills. Through pre-experimental research design, 12 fifth grade SMID participated in the research. To collect data, the researchers used science process skills development forms and rubrics to score their observations. During the teaching intervention, SMID implemented three POE worksheets, which included buzz 22 technique in the ‘prediction’ stage, hands-on experiments and QR codes in the ‘observation’ stage and snowball, learning gallery or card showing techniques in the ‘evaluation’ stage. The results indicated that the POE worksheets are effective at developing SMID’s science process skills of the "matter and its nature" subject. This research recommends that future comparative research should unveil how the POE worksheets with/without active learning techniques impact the SMID’s SPS. Keywords: mild intellectual disabilities, science education, science process skills

Teachers’ Perceptions of Active Learning Techniques in Teaching Speaking: A Case at Language Centers in the Mekong Delta

This paper reports a descriptive study that explored teachers’ perceptions of implementing active learning techniques in teaching speaking and the challenges teachers encountered in their teaching practices in their speaking classes. Participants were 56 teachers from different language centers in the Mekong Delta, Vietnam. The data were collected through questionnaires and semi-structured interviews. The findings of the study revealed that the majority of the teachers had positive perceptions of the implementation of active learning techniques in teaching speaking. They also expressed the advantages they gained as well as the difficulties they faced when implementing active learning techniques in their speaking classes. Keywords: perceptions, active learning, speaking skills DOI: 10.7176/JEP/14-10-02 Publication date: April 30 th 2023

Strategies For Primary School Students Understanding Of Character Education Through The Active Role Of Teachers

Character education plays a vital role in the development of children's mature personalities. The lack of adequate character education strengthening in schools is an issue that requires attention. As facilitators, companions, and sources in the process of conveying knowledge to students about the importance of character education in elementary school, teachers must play a crucial role in the development of children's character. This study aims to analyze the role of teachers in assisting elementary schools in implementing strategies for improving elementary school students' understanding of character education in order to provide benefits in terms of understanding the obstacles in the application of character education in elementary schools and strategies that can be implemented, particularly for teacher-student communication. Using a literature review, this study determines the elements and tactics that can be implemented in primary schools to improve character education for pupils. Recommendations were made for schools based on the findings, including employing active learning techniques and incorporating character education values into all learning materials. Reflection and evaluation should also be conducted to assist students in comprehending and improving their character education-related behavior. To reach the true aims of character education, it is necessary to overcome obstacles such as schools' lack of preparedness, educators' lack of comprehension, and ineffective learning techniques.

The Impacts of Quality-Oriented Dataset Labeling on Tree Cover Segmentation Using U-Net: A Case Study in WorldView-3 Imagery

Deep learning has emerged as a prominent technique for extracting vegetation information from high-resolution satellite imagery. However, less attention has been paid to the quality of dataset labeling as compared to research into networks and models, despite data quality consistently having a high impact on final accuracies. In this work, we trained a U-Net model for tree cover segmentation in 30 cm WorldView-3 imagery and assessed the impact of training data quality on segmentation accuracy. We produced two reference tree cover masks of different qualities by labeling images accurately or roughly and trained the model on a combination of both, with varying proportions. Our results show that models trained with accurately delineated masks achieved higher accuracy (88.06%) than models trained on masks that were only roughly delineated (81.13%). When combining the accurately and roughly delineated masks at varying proportions, we found that the segmentation accuracy increased with the proportion of accurately delineated masks. Furthermore, we applied semisupervised active learning techniques to identify an efficient strategy for selecting images for labeling. This showed that semisupervised active learning saved nearly 50% of the labeling cost when applied to accurate masks, while maintaining high accuracy (88.07%). Our study suggests that accurate mask delineation and semisupervised active learning are essential for efficiently generating training datasets in the context of tree cover segmentation from high-resolution satellite imagery.

Uncertainty-aware predictions of molecular x-ray absorption spectra using neural network ensembles

As machine learning (ML) methods continue to be applied to a broad scope of problems in the physical sciences, uncertainty quantification is becoming correspondingly more important for their robust application. Uncertainty-aware machine learning methods have been used in select applications, but largely for scalar properties. In this work, we showcase an exemplary study in which neural network ensembles are used to predict the x-ray absorption spectra of small molecules, as well as their pointwise uncertainty, from local atomic environments. The performance of the resulting surrogate clearly demonstrates quantitative correlation between errors relative to ground truth and the predicted uncertainty estimates. Significantly, the model provides an upper bound on the expected error. Specifically, an important quality of this uncertainty-aware model is that it can indicate when the model is predicting on out-of-sample data. This allows for its integration with large-scale sampling of structures together with active learning or other techniques for structure refinement. Additionally, our models can be generalized to larger molecules than those used for training, and also successfully track uncertainty due to random distortions in test molecules. While we demonstrate this workflow on a specific example, ensemble learning is completely general. We believe it could have significant impact on ML-enabled forward modeling of a broad array of molecular and materials properties.

A hidden benefit in the COVID-19 pandemic: Rethinking physical geography pedagogy in higher education using a flipped classroom approach

A shift in geographical pedagogy has occurred in the last decade with greater integration of active and blended (hybrid) learning techniques in higher education. A ‘flipped classroom’ approach employs an active classroom and inclusion of pre/post asynchronous course content in replacement of a traditional lecture-based format. This pedagogical research paper investigates the effectiveness of an online flipped classroom approach in the field of physical geography through student evaluation and reflects upon the enforced transition to online learning during the COVID-19 pandemic. An anonymous survey was distributed to final year physical geography students who experienced both on-campus and online education, to evaluate perceived technological skills, broadband access and module delivery preference. Positive qualitative and quantitative results (66 respondents) showed an appreciation for an active classroom, increased student-staff engagement and the use of multimedia resources (i.e. using more than one medium of communication). Sixty-nine per cent of students in this cohort preferred the online flipped classroom to other online lecture-based formats (synchronous or asynchronous). Additionally, 59% of students chose an on-campus flipped classroom approach in replacement of a traditional transmissive lecture or option to remain online in a post-COVID-19 scenario. Overall, this case study provided valuable insight into student experiences of online learning and illuminates a potential pathway to move away from traditional lecture-based methods in undergraduate physical geography. The dynamic, interactive learning environment created by this alternative approach is recommended to enhance learning, accessibility and improve student-staff interaction, particularly for large class sizes where fieldwork may not be feasible.

Five Lessons for Effectively Transitioning Problem-Based Learning to Online Delivery

Problem-based learning (PBL) is an active learning technique that promotes a life-long learning approach to understanding and using the principles of clinical medicine. It does so by helping learners hone their critical thinking skills in a team-based environment. It was originally developed for use in live, in-person settings. During the COVID-19 pandemic, it has had to be rapidly adapted for online delivery. In this article, we first highlight the key challenges faced by educators and learners in making this transition. We then share five lessons for effectively translating in-person PBL curricula to online and hybrid learning formats.

A Survey on Active Learning: State-of-the-Art, Practical Challenges and Research Directions

Despite the availability and ease of collecting a large amount of free, unlabeled data, the expensive and time-consuming labeling process is still an obstacle to labeling a sufficient amount of training data, which is essential for building supervised learning models. Here, with low labeling cost, the active learning (AL) technique could be a solution, whereby a few, high-quality data points are queried by searching for the most informative and representative points within the instance space. This strategy ensures high generalizability across the space and improves classification performance on data we have never seen before. In this paper, we provide a survey of recent studies on active learning in the context of classification. This survey starts with an introduction to the theoretical background of the AL technique, AL scenarios, AL components supported with visual explanations, and illustrative examples to explain how AL simply works and the benefits of using AL. In addition to an overview of the query strategies for the classification scenarios, this survey provides a high-level summary to explain various practical challenges with AL in real-world settings; it also explains how AL can be combined with various research areas. Finally, the most commonly used AL software packages and experimental evaluation metrics with AL are also discussed.

Million-scale data integrated deep neural network for phonon properties of heuslers spanning the periodic table

Existing machine learning potentials for predicting phonon properties of crystals are typically limited on a material-to-material basis, primarily due to the exponential scaling of model complexity with the number of atomic species. We address this bottleneck with the developed Elemental Spatial Density Neural Network Force Field, namely Elemental-SDNNFF. The effectiveness and precision of our Elemental-SDNNFF approach are demonstrated on 11,866 full, half, and quaternary Heusler structures spanning 55 elements in the periodic table by prediction of complete phonon properties. Self-improvement schemes including active learning and data augmentation techniques provide an abundant 9.4 million atomic data for training. Deep insight into predicted ultralow lattice thermal conductivity (<1 Wm−1 K−1) of 774 Heusler structures is gained by p–d orbital hybridization analysis. Additionally, a class of two-band charge-2 Weyl points, referred to as “double Weyl points”, are found in 68% and 87% of 1662 half and 1550 quaternary Heuslers, respectively.

액션러닝 기반 한국어 유사 문법 교사 교육 프로그램 개발 기초 연구

The purpose of this paper is to provide basic data necessary to develop a teacher education program to strengthen similar grammar capabilities of Korean language teachers of less than 5 years by applying Action Learning techniques. To achieve this purpose, the core competencies of similar grammar teachers were presented, and the suitability of behavioral learning and behavioral learning procedures were derived to set the direction of development of the program. Based on the development direction, a draft of a teacher education program in the area of similar grammar was prepared, and the final draft was confirmed through expert review. The procedure of Action Learning was constructed in four stages, and similar grammar teacher education programs and examples were presented. In addition, for the completeness of this study, the implementation of the program proposal and the applicability of other areas other than grammar were proposed as follow-up studies.

Four Ways to Add Active Learning to Computing Courses

How active-learning techniques can benefit students in computing courses.

Promotion of Student Engagement Through the Application of Good Practices in Nursing Online Education.

The purpose of this article is to review Chickering and Gamson's principles of good practice in teaching and to illustrate their applicability to nursing online education delivery. An additional purpose is to present examples of teaching methods used by faculty to promote engagement in online education courses during the pandemic. The original 7 best practices in education, including ( a ) encourages contact between students and faculty, ( b ) develops reciprocity and cooperation among students, ( c ) uses active learning techniques, ( d ) gives prompt feedback, ( e ) emphasizes time on task, ( f ) communicates high expectations, and ( g ) respects diverse talents and ways of learning, remain evidence-based guidelines today. The authors recommend the addition of 2 new best practices: ( a ) incorporating assignment flexibility to meet student learning preferences; and ( b ) applying learning to real-life situations. Having evidence-based guidelines for supporting the role of a teacher in the online learning setting is of paramount importance.