Learning Domain Research Articles

A holistic approach to software fault prediction with dynamic classification

Software Fault Prediction is a critical domain in machine learning aimed at pre-emptively identifying and mitigating software faults. This study addresses challenges related to imbalanced datasets and feature selection, significantly enhancing the effectiveness of fault prediction models. We mitigate class imbalance in the Unified Dataset using the Random-Over Sampling technique, resulting in superior accuracy for minority-class predictions. Additionally, we employ the innovative Ant-Colony Optimization algorithm (ACO) for feature selection, extracting pertinent features to amplify model performance. Recognizing the limitations of individual machine learning models, we introduce the Dynamic Classifier, a ground-breaking ensemble that combines predictions from multiple algorithms, elevating fault prediction precision. Model parameters are fine-tuned using the Grid-Search Method, achieving an accuracy of 94.129% and superior overall performance compared to random forest, decision tree and other standard machine learning algorithms. The core contribution of this study lies in the comparative analysis, pitting our Dynamic Classifier against Standard Algorithms using diverse performance metrics. The results unequivocally establish the Dynamic Classifier as a frontrunner, highlighting its prowess in fault prediction. In conclusion, this research introduces a comprehensive and innovative approach to software fault prediction. It pioneers the resolution of class imbalance, employs cutting-edge feature selection, and introduces dynamic ensemble classifiers. The proposed methodology, showcasing a significant advancement in performance over existing methods, illuminates the path toward developing more accurate and efficient fault prediction models.

A comprehensive review of model compression techniques in machine learning

This paper critically examines model compression techniques within the machine learning (ML) domain, emphasizing their role in enhancing model efficiency for deployment in resource-constrained environments, such as mobile devices, edge computing, and Internet of Things (IoT) systems. By systematically exploring compression techniques and lightweight design architectures, it is provided a comprehensive understanding of their operational contexts and effectiveness. The synthesis of these strategies reveals a dynamic interplay between model performance and computational demand, highlighting the balance required for optimal application. As machine learning (ML) models grow increasingly complex and data-intensive, the demand for computational resources and memory has surged accordingly. This escalation presents significant challenges for the deployment of artificial intelligence (AI) systems in real-world applications, particularly where hardware capabilities are limited. Therefore, model compression techniques are not merely advantageous but essential for ensuring that these models can be utilized across various domains, maintaining high performance without prohibitive resource requirements. Furthermore, this review underscores the importance of model compression in sustainable artificial intelligence (AI) development. The introduction of hybrid methods, which combine multiple compression techniques, promises to deliver superior performance and efficiency. Additionally, the development of intelligent frameworks capable of selecting the most appropriate compression strategy based on specific application needs is crucial for advancing the field. The practical examples and engineering applications discussed demonstrate the real-world impact of these techniques. By optimizing the balance between model complexity and computational efficiency, model compression ensures that the advancements in AI technology remain sustainable and widely applicable. This comprehensive review thus contributes to the academic discourse and guides innovative solutions for efficient and responsible machine learning practices, paving the way for future advancements in the field.Graphical abstract

When Play Reveals the Ache: Introducing Co-constructive Patient Simulation for Narrative Practitioners in Medical Education.

Despite the ubiquity of healthcare simulation and the humanities in medical education, the two domains of learning remain unintegrated. The stories suffused within healthcare simulation have thus remained unshaped by the developments of narrative medicine and the health humanities. Healthcare simulation, in turn, has yet to utilize concepts like co-construction and narrative competence to enrich learners' understanding of patient experience alongside their clinical competencies. To create a conceptual bridge between these two fields (including narrative-based inquiry more broadly), we redescribe narrative competence via Ronald Heifetz's distinction of "technical" and "adaptive" challenges outlined in his adaptive leadership model. Heifetz, we argue, enriches learners' self-understanding of the unique demands of cultivating narrative competence, which can be both elucidated on the page and tested within the charged yet supportive simulation environment. We introduce Co-constructive Patient Simulation (CCPS) to demonstrate how working with simulated patients can support narrative work by drawing on the clinical vicissitudes of learners in the formulation and enactment of case studies. The three movements of CCPS-resensing, retelling, and retooling-told through learner experiences, describe the affinities and divergences between narrative medicine's sequence of attention, representation, and affiliation; Montello's three forms of narrative competence (departure, performance, change), and Heifetz's three steps (observe, interpret, and intervene) of adaptive leadership.

Clinical research self-efficacy in academic veterinary clinicians in the United States.

Clinical research is a growing part of the academic clinician's job, and documenting areas of low self-efficacy can inform training initiatives. 182 US academic veterinary clinicians. A survey of academic veterinary clinicians was distributed to 31 US institutions. Self-efficacy was assessed with a modified Clinical Research Appraisal Instrument-12. The relationship between research self-efficacy and completion of formal research training and years of experience was evaluated. Respondents were predominantly junior to midcareer faculty. The lowest reported confidence was in performing advanced statistical analyses (3; 0 to 10). Other low-confidence tasks included designing a qualitative methods study (4; 0 to 10), terminating a collaboration that isn't working (5; 0 to 10), describing a funding agency's review/award process (5; 0 to 10), establishing a timeline for a grant application (5; 0 to 10), establishing collaborator and consultant agreements (5; 0 to 10), and asking staff to leave the project team (5; 0 to 10). Completion of a formal research training program was significantly associated with improved self-efficacy in many tasks. Years of experience was also associated, especially in project management and interpersonal interactions. These results highlight the need for targeted training opportunities for academic veterinary clinicians in biostatistical support, qualitative study design methods, and aspects of communication and interpersonal skills that are important for developing and leading effective research teams. Range of confidence suggests that development opportunities in all domains will improve self-efficacy for some clinicians. Future studies should focus on the impact of formal training sessions on various domains of self-efficacy and on targeted mentoring in supporting confidence in more experiential learning domains.

Continuing Medical Education (CME) on the Family Adoption Program

The Family Adoption Program has been incorporated into the MBBS curriculum starting from the 2021-22 batch. This change in the curriculum aims to align it with competency-based medical education and ensure comprehensive coverage of cognitive, affective, and psychomotor domains of learning. The objective of the program is to enhance the accessibility of healthcare services to the rural and underserved people. This is necessary because the existing healthcare facilities in our country are primarily located in urban regions, while the rural areas have a higher concentration of population but limited health care facilities as well as poor access to healthcare. [1] A two-hour session was conducted under the banner of pre-conference CME of JHIAPSMCON 2024 on the topic- Family Adoption Program: A roller coaster ride. There were about 30 participants representing nine medical colleges of Jharkhand. The session was chaired by two faculty from an autonomous Government institution of the state. The agenda for discussion were the brief introduction on Family Adoption Program, its implementation, the role of health camp in FAP and ways in which students may be engaged in health camp and lastly the strengths, weaknesses, opportunities and challenges of the program. The intention was to share the experience of FAP implementation at Manipal Tata Medical College and also exchange of ideas and experiences by all other institutions. The perspective of faculties from other institution were also discussed. [2] There were four resource person for the session which continued one after the other with a brief discussion with the participants at the end of each sub-session.

Neurocognitive Functioning is Impaired in Perinatally HIV-Infected Youth.

The present study examined neurocognitive differences between Perinatally HIV (PHIV)-infected-youth and age and gender matched healthy controls. Despite early, long-term anti-viral treatment (ART), significant neurocognitive deficiencies remain for PHIV-infected-youth reaching adulthood compared to controls. Participants were assessed with a comprehensive neuropsychological battery. An Overall Neurocognitive Composite Score and a Global Deficit Score (GDS) were created. Sleep, depression, and developmental level of intellectual functioning were also examined. PHIV-youth performed more poorly than controls in all neurocognitive domains. Very large effect sizes were observed for the Overall Neurocognitive Composite Score and GDS. PHIV-infected-youth appear to be significantly more depressed compared to controls, but there were no differences in amount or type of sleep observed. Despite early, long-term anti-viral treatment (ART), neurocognitive deficiencies remain for PHIV-infected-young-adults. The verbal learning domain was significantly impaired with implications for functioning. The PHIV-infected-youth were also depressed and not receiving treatment for depression.

Contrastive Learning with Transformer Initialization and Clustering Prior for Text Representation

Acquiring labeled data for learning sentence embeddings in Natural Language Processing poses challenges due to limited availability and high costs. In order to tackle this issue, we introduce a novel method called Contrastive Learning with Transformer Initialization and Clustering Prior for Text Representation (CLTC). Our method utilizes Pre-Layernorm Transformers without warm-up, stabilizing the training process while also increasing the final performance. We employ Contrastive Learning (CL) with dropout-based augmentation to enhance sentence embeddings. Additionally, we integrate prior knowledge into the contrastive learning framework within an efficient clustering strategy. When evaluated on the SentEval task, our approach showcases a competitive performance when compared to state-of-the-art approaches in the contrastive learning domain. Our method offers stability, improved embeddings, and the utilization of prior knowledge for enhanced unsupervised representation learning in Natural Language Processing applications.

Adaptive spatiotemporal neural networks through complementary hybridization

Processing spatiotemporal data sources with both high spatial dimension and rich temporal information is a ubiquitous need in machine intelligence. Recurrent neural networks in the machine learning domain and bio-inspired spiking neural networks in the neuromorphic computing domain are two promising candidate models for dealing with spatiotemporal data via extrinsic dynamics and intrinsic dynamics, respectively. Nevertheless, these networks have disparate modeling paradigms, which leads to different performance results, making it hard for them to cover diverse data sources and performance requirements in practice. Constructing a unified modeling framework that can effectively and adaptively process variable spatiotemporal data in different situations remains quite challenging. In this work, we propose hybrid spatiotemporal neural networks created by combining the recurrent neural networks and spiking neural networks under a unified surrogate gradient learning framework and a Hessian-aware neuron selection method. By flexibly tuning the ratio between two types of neurons, the hybrid model demonstrates better adaptive ability in balancing different performance metrics, including accuracy, robustness, and efficiency on several typical benchmarks, and generally outperforms conventional single-paradigm recurrent neural networks and spiking neural networks. Furthermore, we evidence the great potential of the proposed network with a robotic task in varying environments. With our proof of concept, the proposed hybrid model provides a generic modeling route to process spatiotemporal data sources in the open world.

Graph sequence learning for premise selection

Premise selection is crucial for large theory reasoning with automated theorem provers as the sheer size of the problems quickly leads to resource exhaustion. This paper proposes a premise selection method inspired by the machine learning domain of image captioning, where language models automatically generate a suitable caption for a given image. Likewise, we attempt to generate the sequence of axioms required to construct the proof of a given conjecture. In our axiom captioning approach, a pre-trained graph neural network is combined with a language model via transfer learning to encapsulate both the inter-axiom and conjecture-axiom relationships. We evaluate different configurations of our method and experience a 14% improvement in the number of solved problems over a baseline.

Utilizing biological experimental data and molecular dynamics for the classification of mutational hotspots through machine learning.

Benzo[a]pyrene, a notorious DNA-damaging carcinogen, belongs to the family of polycyclic aromatic hydrocarbons commonly found in tobacco smoke. Surprisingly, nucleotide excision repair (NER) machinery exhibits inefficiency in recognizing specific bulky DNA adducts including Benzo[a]pyrene Diol-Epoxide (BPDE), a Benzo[a]pyrene metabolite. While sequence context is emerging as the leading factor linking the inadequate NER response to BPDE adducts, the precise structural attributes governing these disparities remain inadequately understood. We therefore combined the domains of molecular dynamics and machine learning to conduct a comprehensive assessment of helical distortion caused by BPDE-Guanine adducts in multiple gene contexts. Specifically, we implemented a dual approach involving a random forest classification-based analysis and subsequent feature selection to identify precise topological features that may distinguish adduct sites of variable repair capacity. Our models were trained using helical data extracted from duplexes representing both BPDE hotspot and nonhotspot sites within the TP53 gene, then applied to sites within TP53, cII, and lacZ genes. We show our optimized model consistently achieved exceptional performance, with accuracy, precision, and f1 scores exceeding 91%. Our feature selection approach uncovered that discernible variance in regional base pair rotation played a pivotal role in informing the decisions of our model. Notably, these disparities were highly conserved among TP53 and lacZ duplexes and appeared to be influenced by the regional GC content. As such, our findings suggest that there are indeed conserved topological features distinguishing hotspots and nonhotpot sites, highlighting regional GC content as a potential biomarker for mutation. Code for comparing machine learning classifiers and evaluating their performance is available at https://github.com/jdavies24/ML-Classifier-Comparison, and code for analysing DNA structure with Curves+ and Canal using Random Forest is available at https://github.com/jdavies24/ML-classification-of-DNA-trajectories.

A systematic review of genomic education for nurses and nursing students: Are they sufficient in the era of precision health?

BackgroundIntegration of genomic content into nursing curricula and continuing education is limited. PurposeThis systematic review aimed to identify educational strategies to deliver genomics content. MethodCINAHL Plus, ERIC, Ovid MEDLINE, and Scopus electronic databases were searched from January 1, 2003 through July 6, 2023. The three domains of learning: cognitive, affective, and psychomotor were used to categorize educational outcomes. A narrative approach was used to synthesize the data using Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines. FindingsTen studies met inclusion criteria. Quality assessment for bias was conducted independently; study quality ranged from fair to poor. No study used a complete psychometrically tested instrument to measure genomic cognitive or affective knowledge; none measured the psychomotor domain. DiscussionEffective strategies to deliver genomic education is a priority. Development of reliable and valid instruments for the assessment of educational interventions including the psychomotor skills needed for practice.

Embracing imperfections: a predictive analysis of factors alleviating adult learners’ digital learning stress on Singapore’s lifelong learning journey

ABSTRACT This research paper examines the intersection of self-determination theory and its constructs of autonomy, competence and relatedness with the effects of learning support levers available to online educators. The learning support levers of social, technical and financial support are considered against their ability to impact online learning satisfaction and overall digital well-being. In surveying 246 adult learners, the research analysis uses structural equation modelling to yield results that highlight the opportunity for suitably planned social and technical support design to enhance the effectiveness of online teaching and learning. The overall findings add to understanding in the domain of digital learning effectiveness and possible strategies for improving online teaching and learning instructional design, including the surrounding environmental support variables. The paper further highlights the benefits of a cross domain approach to exploring learning effectiveness goals in a digital context. Given the increased prevalence of online education, the methodology and resultant findings of this paper are novel, reflecting the lack of research in the digital learning domain.

Minds in movement: embodied cognition in the age of artificial intelligence.

This theme issue brings together researchers from diverse fields to assess the current status and future prospects of embodied cognition in the age of generative artificial intelligence. In this introduction, we first clarify our view of embodiment as a potentially unifying concept in the study of cognition, characterizing this as a perspective that questions mind-body dualism and recognizes a profound continuity between sensorimotor action in the world and more abstract forms of cognition. We then consider how this unifying concept is developed and elaborated by the other contributions to this issue, identifying the following two key themes: (i) the role of language in cognition and its entanglement with the body and (ii) bodily mechanisms of interpersonal perception and alignment across the domains of social affiliation, teaching and learning. On balance, we consider that embodied approaches to the study of cognition, culture and evolution remain promising, but will require greater integration across disciplines to fully realize their potential. We conclude by suggesting that researchers will need to be ready and able to meet the various methodological, theoretical and practical challenges this will entail and remain open to encountering markedly different viewpoints about how and why embodiment matters. This article is the part of this theme issue 'Minds in movement: embodied cognition in the age of artificial intelligence'.

Decentralized Counterfactual Value with Threat Detection for Multi-Agent Reinforcement Learning in mixed cooperative and competitive environments

This paper proposes a fully decentralized approach to address the challenge of general mixed cooperation and competition within the domain of Multi-Agent Reinforcement Learning (MARL). Conventional MARL approaches do not achieve full decentralization as they necessitate either the communication of implicit information or the retention of a centralized critic, rendering them impractical in mixed cooperative and competitive environments. To address these challenges, this paper proposes a Decentralized Counterfactual Value (DCV) to model the behaviors of other agents and mitigate the non-stationary problem, accompanied by a Threat Detection (TD) mechanism to discern latent competitive or cooperative relationships. In addition, DCVTD is incorporated into both value-based and policy-based RL paradigms with theoretical convergence guarantee. Finally, empirical validation across four representative environments demonstrates the superior performance of DCVTD in terms of collective returns, computational efficiency, and agent scalability over other fully decentralized approaches, centralized training with decentralized execution approaches, and alternative approaches involving agent modeling or reward shaping in comprehensive experiments.

Deep dive into RNA: a systematic literature review on RNA structure prediction using machine learning methods

The discovery of non-coding RNAs (ncRNAs) has expanded our comprehension of RNAs’ inherent nature and capabilities. The intricate three-dimensional structures assumed by RNAs dictate their specific functions and molecular interactions. However, the limited number of mapped structures, partly due to experimental constraints of methods such as nuclear magnetic resonance (NMR), highlights the importance of in silico prediction solutions. This is particularly crucial in potential applications in therapeutic drug discovery. In this context, machine learning (ML) methods have emerged as prominent candidates, having previously demonstrated prowess in solving complex challenges across various domains. This review focuses on analyzing the development of ML-based solutions for RNA structure prediction, specifically oriented toward recent advancements in the deep learning (DL) domain. A systematic analysis of 33 works reveals insights into the representation of RNA structures, secondary structure motifs, and tertiary interactions. The review highlights current trends in ML methods used for RNA structure prediction, demonstrates the growing research involvement in this field, and summarizes the most valuable findings.

Multi-modal data clustering using deep learning: A systematic review

Multi-modal clustering represents a formidable challenge in the domain of unsupervised learning. The objective of multi-modal clustering is to categorize data collected from diverse modalities, such as audio, visual, and textual sources, into distinct clusters. These clustering techniques operate by extracting shared features across modalities in an unsupervised manner, where the identified common features exhibit high correlations within real-world objects. Recognizing the importance of perceiving the correlated nature of these features is vital for enhancing clustering accuracy in multi-modal settings. This survey explores Deep Learning (DL) techniques applied to multi-modal clustering, encompassing methodologies such as Convolutional Neural Networks (CNN), Autoencoders (AE), Recurrent Neural Networks (RNN), and Graph Convolutional Networks (GCN). Notably, this survey represents the first attempt to investigate DL techniques specifically for multi-modal clustering. The survey presents a novel taxonomy for DL-based multi-modal clustering, conducts a comparative analysis of various multi-modal clustering approaches, and deliberates on the datasets employed in the evaluation process. Additionally, the survey identifies research gaps within the realm of multi-modal clustering, offering insights into potential future avenues for research.

BDSM Safety in Pornography: Its Perceived Accuracy and Impact on Practitioners' Safe Word Practices.

As some BDSM activities can carry risk of temporary or permanent marks/injury, and even death, safety is a foundational principle to BDSM participation. Given the connection between BDSM and sex for some practitioners, pornography is a likely domain for learning about BDSM safety, given its easy accessibility online. However, it is unclear whether practitioners view BDSM safety depicted in pornography as accurate and whether it influences their own practices. Therefore, the current study surveyed self-identified BDSM practitioners' perceptions of pornography's accurate representation of communication, consent, safe words, negotiation, general safety, and how these perceptions relate to their use of safe words and safety training in their own activities. Results showed that (1) demographics nor BDSM-related attributes predicted perceived accuracy of BDSM safety depictions in pornography; (2) perceived accuracy of pornography's depiction of BDSM general safety and safe words did not predict practitioners' negotiation or use of safe words; however, years of BDSM experience did predict safe word usage; (3) submissive-related BDSM roles were more likely to use safe words than dominant-related BDSM roles; (4) those who participated in BDSM only in private were less likely to have CPR and/or first-aid training and more likely to have no other safety training; (5) those who believed pornography depicted BDSM safety mostly/very accurately were more likely to have no safety training; and (6) years of BDSM experience did not predict safety training. The impact of these findings on how we understand the influence of pornography on real-world BDSM practices are discussed.

Assessment of the medical school learning environment at United Arab Emirates University

BackgroundThe purpose of the article was to assess students’ perception of the learning environment at the College of Medicine and Health Sciences, United Arab Emirates University using the ‘Medical School Learning Environment Survey’ (MSLES). Evaluating the learning environment and working towards its improvement is crucial for the physical and mental well-being of medical students, as it contributes to fostering an optimal learning environment.MethodsStudents participated in four groups: Year-1 (pre-medical), Year-2 (pre-medical), Year-3/Year-4 (pre-clinical), and Year-5/Year-6 (clinical). MSLES data was collected from each group using an online survey tool (Qualtrics XM). Latent factor structures in the learning environment constructs were assessed using ‘exploratory factor analysis’; and reliability was measured by Cronbach’s alpha. Differences among groups were assessed by ‘single factor ANOVA’.ResultsThree hundred seventy-seven (65%) of the 584 eligible students completed the survey. Exploratory factor analysis revealed four factors: (Genn J. AMEE Medical Education Guide 23 (Part 1): Curriculum, environment, climate, quality and change in medical education–a unifying perspective. Med Teach. 2001;23(4):337–44.) Learning experience (∝=0.71), (Maudsley RF. Role models and the learning environment: essential elements in effective medical education. Acad Med. 2001;76(5):432–4.) Student-student interaction (∝=0.69), (Gruppen LD, Irby DM, Durning SJ, Maggio LA. Conceptualizing learning environments in the health professions. Acad Med. 2019;94(7):969–74.) Student-faculty interaction (∝=0.62), and (Soemantri D, Herrera C, Riquelme A. Measuring the educational environment in health professions studies: A systematic review. Med Teach. 2010;32(12):947–52.) Academic support (∝=0.62). Students in Y3-Y6 rated the learning environment statistically significantly lower than that in Y1-Y2. Student-student interaction in Y2 was significantly lower than that in other years. Student-faculty interaction in Y1 was significantly higher than that in Y2. Academic support was significantly higher in Y1 than that in Y2-Y6.ConclusionThe MSLES revealed variabilities in learning domains across the years. Improvement efforts should foster student-student collaboration in Y2 and improve academic support approaches for Y2-6. These findings provide valuable insights for medical educators to enhance the medical school learning environment and foster an optimal learning environment in lifelong medical education.

Longitudinal Study of Cognitive Function in People with HIV and Toxoplasmic Encephalitis or Latent toxoplasma Infection.

Neurocognitive impairment (NCI) may occur during and persist even after recovery from HIV-related CNS co-infections such as toxoplasmic encephalitis (TE). The long-term cognitive effects of TE and latent toxoplomasmic infections (LTI) among persons with HIV (PWH) are unknown. We measured longitudinal effects on NC functioning in PWH with TE compared to LTI or no toxoplasmal infection. PWH (n = 345) followed in two longitudinal cohort studies underwent comprehensive neurocognitive assessments and an anti-Toxoplamic IgG assay. Participants were classified into one of three groups: TE+ (n = 39), LTI+ (n = 34), LTI- (n = 272). The primary outcome was change in neurocognitive function between baseline and 7-year visit. The mean age was 48 ± 11 years, mean educational level 13 ± 3 years, and 13% were female. TE+ patients were less likely to have undetectable viral loads (≤50 copies/mL) and had lower absolute CD4 counts. The TE+ group had the highest prevalence of NCI globally and in domains of verbal, executive function, learning, recall, working memory, processing speed and motor at baseline and at 7-year follow-up. Changes in longitudinal NC function over 7 years were small and did not differ significantly among all groups, except that speed of information processing improved more in TE+ compared with LTI- participants. PWH with a history of TE had cognitive impairment over a broad range of severity at both baseline and last follow-up. Changes in cognition from baseline to last examination in all groups were minimal and did not differ significantly among the groups with the exception of speed of information processing.

Beyond content delivery: harnessing emotional intelligence for community building in fully online digital spaces

The onset of the pandemic catalyzed a paradigm shift in educational methodologies, bringing various forms, such as hybrid, distance, and fully online models, into focus. The following study explores the affective domain in online learning, focusing on how emotions, facial expressions, and body language influence engagement and support community building in fully online learning environments. This research explores the role of emotional intelligence in Fully Online Learning Communities (FOLC) and examines the impact of positive and negative emotions on interpersonal engagement and participation. Findings indicate positive emotions to be closely linked to increased engagement and active participation. The study also highlights the importance of exploring body language in digital learning environments and addresses challenges posed by technological barriers in fully online learning spaces. Emotional intelligence is pivotal in online learning and community building, emphasizing the need to understand how to create emotionally supportive digital learning environments. Outcomes indicate a need for future research to focus on understanding the role of cultural dimensions in supporting learner agency and community building in the fully online learning context.