Scale Problems Research Articles

Scheduling analysis of automotive glass manufacturing systems subject to sequence-independent setup time, no-idle machines, and permissive maximum total tardiness constraint

With the increasing demand for automotive glass, improving the efficiency of automotive glass manufacturing systems can make full use of production resources and reduce the waste of natural and social resources. Therefore, this work aims to provide an efficient method for a real-world two-stage hybrid flow shop scheduling problem with small batches in an automotive glass manufacturing system. For the investigated problem, there is a significant setup time at the first stage, the second stage is served by machines that should not be interrupted, and each batch has a due date. Such constraints make this scheduling problem challenging. To solve this problem, a mixed integer linear programming model is established. Also, two properties and three theorems are given to enhance the problem-solving process. Subsequently, an efficient genetic algorithm is carefully designed to solve large-scale problems by considering the properties of the system. Meanwhile, an improvement scheme is proposed to decrease the running time of the algorithm, and experimental results show that this scheme can reduce the running time by 280 s at most from the average results of different scale problems. Finally, extensive experiments are carried out and a real-world case is solved to demonstrate the efficiency and effectiveness of the designed genetic algorithm. Also, the Taguchi method is adopted to tune the parameters for the designed algorithm.

Emotional and behavioural problems of left behind children in Lithuania: a comparative analysis of youth self-reports and parent/caregiver reports using ASEBA.

Children being left behind (LBC) in their home countries due to parental emigration is a global issue. Research shows that parents' emigration negatively affects children's mental health and well-being. Despite a high number of LBC, there is a dearth of data from Eastern European countries. The present study aims to collect and analyse self-reported data on LBC emotional and behavioural problems and compare children's reports with those of parents/caregivers. A cross-sectional study was conducted in 24 Lithuanian schools, involving parents/caregivers and their children aged 12 to 17. We employed self-reported measures, including the Achenbach System of Empirically Based Assessment (ASEBA) tools -Child Behaviour Checklist (CBCL 6/18) and Youth Self Report (YSR 11/18), to evaluate the emotional and behavioural problems of the children. These instruments had been translated, standardised, and validated for the Lithuanian population. Data collection took place between January 2022 and April 2023. In addition to descriptive analysis, multivariate regression was used to adjust for various sociodemographic factors. A total of 760 parents/caregivers and 728 of their children participated in the study. LBC exhibited higher total problem scores (57.7; 95% CI 52.0-63.4) compared to non-LBC (47.1; 95% CI 44.7-49.4). These differences were consistent across all YSR 11/18 problem scales. However, no significant differences were observed in CBCL 6/18 scores. Furthermore, LBC self-reported a higher total problem score (57.7; 95% CI 52.0-63.4) compared to their parents/caregivers (24.9; 95% CI 18.9-30.9), and this pattern persisted across all scales. Being female, having school-related problems and having LBC status were associated with higher YSR 11/18 scores in the multivariable regression, while female gender, living in rural areas, school-related problems, and having hobbies were associated with higher CBCL 6/18 scores. This study highlights that LBC report more emotional and behavioural challenges than their non-LBC peers, while parent/caregiver assessments show lower problem scores for LBC. Gender, living environment, school-related issues, and engagement in hobbies have influenced these outcomes. These findings underscore the multifaceted nature of the experiences of LBC and the importance of considering various contextual factors in understanding and addressing their emotional and behavioural well-being.

Associations between Atopic Dermatitis and Behavior Difficulties in Children.

Background and Objectives: There has been increasing evidence that atopic dermatitis (AD) is associated with behavioral difficulties (BDs). There is currently a lack of evidence of how the severity of the disease determines BDs and what additional factors may contribute to their manifestation. The aim is to determine what kind of BDs occur in children with AD compared to healthy children and to find out what additional factors may contribute to the development of BDs in AD patients. Materials and Methods: This is a cross-sectional, prospective study with the application of a risk assessment instrument for behavior difficulties (Child Behavior Checklist, CBCL 6/18) in pediatric patients with AD and healthy controls (6-17 years) between 1 January 2020 and 31 December 2022. For statistical comparison, mainly Wilcoxon-Mann-Whitney and Student's t-test were used, considering a significance level of 5%. Results: This study included a total of 101 children: 48% with AD, 52% non-AD. The mean age was 10 ± 2.7 years for AD, and10.5 ± 3.1 years for the control patients. AD patients had higher internal behavior scale scores and T-scores (6.6 ± 6.4 vs. 9.6 ± 6.9 and 47.9 ± 9.5 vs. 52.3 ± 10.2, p = 0.01), anxiety/depression scale score and T-score (2.8 ± 2.7 vs. 4.3 ± 3.5 and 47.7 ± 8.4 vs. 52.5 ± 11, p = 0.02), and somatic problems scale score and T-score (2.1 ± 2.3 vs. 3.5 ± 3 and 47.6 ± 8.5 vs. 52.7 ± 10.9, p = 0.005). Patients with severe AD had sleep disturbance and itching scores higher than those with mild-moderate AD (5.4 ± 2.6 vs. 2.4 ± 2.2, p = 0.000 and 6.6 ± 2.4 vs. 4 ± 2.8, p = 0.001). The mean morning serum cortisol concentration was lower in AD patients compared to controls (252.91 ± 304.34 vs. 351.55 ± 126.09 nmol/L, p = 0.047). Conclusions: Children with AD present a higher risk of BDs than healthy controls. Patients with severe AD experience more sleep disturbances and a greater intensity of itching compared to mild-moderate AD. The occurrence of BDs was not related to serum cortisol levels. The cortisol level, severity, age, gender, duration of illness, intensity of pruritus, and sleep disturbance did not affect the development of BDs.

Advanced Applications of Artificial Intelligence in Pharmacovigilance: Current Trends and Future Perspectives

The primary goal of pharmacovigilance, the cornerstone of public health, is to track and evaluate adverse drug reactions in order to guarantee patient safety. Conventional approaches suffer from biases in human error, inefficiency, and scalability problems. A new era in pharmacovigilance is being ushered in by the introduction of artificial intelligence (AI), which holds the promise of vast data analysis, automated procedures, and enhanced safety signal detection. Artificial intelligence technologies improve adverse event detection and signal identification by providing unparalleled speed, accuracy, and scalability. They are adept at applying sophisticated algorithms, machine learning models, and natural language processing to glean insights from unstructured data sources like clinical notes, patient narratives, and regulatory reports. This capacity makes it possible to implement risk management plans that are more thorough and proactive. However, using artificial intelligence in pharmacovigilance necessitates large expenditures for processing power, infrastructure, and regulatory compliance. For artificial intelligence-driven systems to be accurate, dependable, and applicable, ongoing validation, monitoring, and improvement efforts are essential. The ethical and legal implications of patient privacy, data security, and regulatory compliance underscore the necessity of cautious artificial intelligence deployment in order to maintain public trust and protect patient rights. Regulatory agencies, healthcare professionals, and artificial intelligence developers must work together efficiently to implement explainable artificial intelligence frameworks, adaptive surveillance techniques, and improved signal validation processes in order to fully realize the potential of artificial intelligence in pharmacovigilance. This will usher in a new era of proactive risk assessment and improved public health outcomes. Keywords Pharmacovigilance, Artificial Intelligence, Signal Detection, Regulatory Affairs, Patient Safety, Public Health

Preliminary thermal performance characterization of a flexible separate heat pipe used for the tracking-type photovoltaic/thermal system

In the solar heating field, the separate heat pipe-based photovoltaic/thermal technology can improve the comprehensive utilization efficiency of solar energy and reduce the maintenance difficulty in long-term outdoor operation by preventing freezing and scaling problems. Its promotion and application are of great significance for reducing fossil energy consumption and carbon emissions, and thus have attracted much attention. However, the performance of conventional photovoltaic/thermal modules is limited by their frame shadow due to the insulation demand, which leaves plenty of room for overall efficiency gains. To solve this problem, a novel separate heat pipe-based photovoltaic/thermal system is proposed by introducing a dual-axis tracking method through a flexible separate heat pipe and side-altitude inclination. To preliminarily characterize the thermal performance of the flexible separate heat pipe in dual-axis tracking conditions, an indoor experimental platform of the flexible separate heat pipe using the solar collector and finned-tube heat exchanger in the water tank as the evaporator and condenser was set up. The effects of side inclination angle, altitude angle of the solar collector and solar irradiance on the thermal efficiency, mass flow rate, and thermal resistance of the system are investigated in detail. The experimental results show the system can generally maintain effective heat transfer in side inclination states, with thermal efficiency mostly ranging from 44.68% to 86.12%. The proposed flexible separate heat pipe is qualified for the dual-axis tracking of photovoltaic/thermal system and shows great potential in improving the utilization of solar energy.

Ruxolitinib Adherence in Myelofibrosis and Polycythemia Vera: the "RAMP" Italian multicenter prospective study.

Ruxolitinib is beneficial in patients with myelofibrosis (MF) and polycythemia vera (PV). Information on ruxolitinib adherence is scant. The Ruxolitinib Adherence in Myelofibrosis and Polycythemia Vera (RAMP) prospective multicenter study (NCT06078319) included 189 ruxolitinib-treated patients. Patients completed the Adherence to Refills and Medications Scale (ARMS) and Distress Thermometer and Problem List (DTPL) at the earliest convenience, after registration in the study, and at later timepoints. At week-0, low adherence (ARMS > 14) and high distress (DT ≥ 4) were declared by 49.7% and 40.2% of patients, respectively. The main reason for low adherence was difficult ruxolitinib supply (49%), intentional (4.3%) and unintentional (46.7%) non-take. In multivariable regression analysis, low adherence was associated to male sex (p = 0.001), high distress (p < 0.001), and treatment duration ≥ 1year (p = 0.03). Over time, rates of low adherence and high distress remained stable, but unintentional non-take decreased from 47.9% to 26.0% at week-48. MF patients with stable high adherence/low distress were more likely to obtain/maintain the spleen response at week-24. Low adherence to ruxolitinib represents an unmet clinical need that require a multifaceted approach, based on reason behind it (patients characteristics and treatment duration). Its recognition may help distinguishing patients who are truly refractory and those in need of therapy optimization.

Quality of life of patients with multiple sclerosis after high-dose immunosuppressive therapy with autologous hematopoietic stem cell transplantation

Introduction. Multiple sclerosis (MS) is a chronic autoimmune and neurodegenerative disease of the central nervous system (CNS). High-dose immunosuppressive therapy with autologous hematopoietic stem cell transplantation (HDIT–AHSCT) is the most eff ective method of treating MS, which has a positive eff ect on the quality of life of patients and inhibition of the progression of disability.The aim of the study was to analyze the dynamics of quality of life in patients with MS who underwent the HDIT– AHSCT method in groups with diff erent clinical effi cacy of this method.Material and methods. The single-center observational study included patients with MS (n = 21, mean age 35.5 ± 6.2 (28–50) years; men/women — 11/10; remitting MS (RMS) n = 16, secondary-progressive (SPMS) n = 4, primaryprogressive (PPMS) n = 1), after HDIT–AHSCT (cyclophosphamide-rituximab conditioning regimen). Assessment of the quality of life (SF-36, FAMS, CSP-MS42, HADS) and the degree of disability in all patients was carried out before and 12 months after HDIT–AHSCT.Results. 12 months after HDIT–AHSCT, more than 50% of patients showed signifi cant improvement in quality of life according to the scales of viability and social functioning of the SF-36 questionnaire, as well as according to the scales of general well-being and additional problems of the FAMS questionnaire. 57% of patients had a signifi cant improvement in the overall FAMS score. According to the CSP-MS42 questionnaire, a decrease in weakness (p &lt; 0.05), feelings of constant fatigue and rapid fatigue during exercise was revealed. After transplantation, there was a slight decrease in anxiety (p = 0.91) and a signifi cant decrease in depression (p = 0.011).Conclusion. The expected dynamics was established: in patients with clinical improvement, there was an improvement in the quality of life for most indicators; in patients with stabilization, most quality of life indicators were similar before and after HDIT–AHSCT; with clinical deterioration after HDIT–AHSCT, quality of life parameters worsened. These data indicate the expediency of further optimization of patient selection criteria and various stages of the HDIT–AHSCT in order to obtain optimal clinical results and, accordingly, improve the quality of life of patients.

Are changes in sleep problems associated with changes in life satisfaction during the retirement transition?

Retirement reduces sleep problems, but changes in life satisfaction during the retirement transition are multifactorial and partly unknown. The aim of this prospective cohort study was to examine whether changes in sleep problems are associated with changes in total and domain-specific life satisfaction during the retirement transition (on average 0.5 years before and 0.5 years after retirement). The study population consisted of Finnish public sector employees (n = 3518) from the Finnish Retirement and Aging (FIREA) study who responded to annual surveys before and after transition to statutory retirement. Sleep problems were measured with Jenkins Sleep Problem Scale questionnaire and participants were grouped into four sleep problem groups depending on the state of their sleep problems during the retirement transition: ‘Never,’ ‘Decreasing,’ ‘Increasing,’ and ‘Constant’ sleep problems. Life satisfaction was measured with the Life Satisfaction Scale questionnaire including four domains (interestingness, happiness, easiness, togetherness). We found that the improvement in total life satisfaction was greatest for participants in the ‘Decreasing’ (0.17, 95% CI 0.11–0.23, SMD 0.27) and ‘Constant’ (0.12, 95% CI 0.07–0.18, SMD 0.19) sleep problem groups. Of the specific life satisfaction domains, similar findings were observed only for the easiness domain. It seems that decreasing or constant sleep problems are associated with improved life satisfaction during the retirement transition, especially in the feeling of easiness of life. This may be due to the fact that as the demands of working life are removed, sleep problems are alleviated or it becomes easier to live with them, which improves life satisfaction.

Impact of Uncertain Flight Time on Heterogeneous UAVs’ Task Planning with Temporal Constraints

Heterogeneous multi-UAV systems offer distinct advantages through their complementary and coordinated use of their diverse capabilities. However, this complexity poses significant challenges in task planning, particularly in considering temporal constraints among tasks. As task dependencies evolve from simple linear chains to complex networked associations, uncertainties in flight times can have a substantial impact on the overall schedule. To address these challenges, this study introduces a rapid estimation method that recursively calculates task completion times, derives their probability distributions, and assesses the robustness of the plan. Furthermore, a neighborhood search algorithm guided by dynamic time windows is designed to effectively evaluate the consequences of task insertions, precisely to adjust high-risk tasks, and reduce blindness in enumerative neighborhood exploration. Simulation results demonstrate that the proposed approach effectively accounts for inherent randomness in the problem and exhibits strong adaptability to changes in the problem scale, flight time fluctuations, and variations in time window constraints.

The Effect of Video-Based Simulation Training on Nursing Students' Motivation and Academic Achievement: A Mixed Study.

This study was conducted to determine the effect of video-based simulation education on nursing students' motivation and academic achievement. The research was designed in a mixed model. A quasi-experimental method with a pretest-posttest control group was used for the quantitative part, and the descriptive phenomenology approach was used as the qualitative research method. The sample of the study consisted of second-year nursing students in two state universities in eastern Turkey. The data were collected with the Student Information Form, the Academic Achievement Test, and the Motivation Resources and Problems Scale using Google Forms Web application. Qualitative data were collected through online semistructured interview forms and focus group interviews. According to the results, the posttest academic achievement and Motivation Resources and Problems Scale mean scores of the students in the intervention group were significantly higher than those of the control group ( P < .05). In the analysis of the qualitative, three main themes emerged: We felt fortunate that it increased information retention," "We felt like we were in real practice environment," and "It made us feel that we were nurses." The results showed the use of video-based simulation can be suggested as a strategy to promote classroom teaching and engage students.

High-Frequency Heart Rate Variability Is Prospectively Associated With Sleep Complaints in a Healthy Working Cohort.

Vagus nerve functioning, as indexed by high-frequency heart rate variability (HF-HRV), has been implicated in a wide range of mental and physical health conditions, including sleep complaints. This study aimed to test associations between HF-HRV measured during sleep (sleep HF-HRV) and subjective sleep complaints 4 years later. One hundred forty-three healthy employees (91% male; MAge = 47.8 years [time 2], SD = 8.3 years) of an industrial company in Southern Germany completed the Jenkins Sleep Problems Scale, participated in a voluntary health assessment, and were given a 24-hour ambulatory heart rate recording device in 2007. Employees returned for a health assessment and completed the Jenkins Sleep Problems Scale 4 years later. Hierarchical regression analyses showed that lower sleep HF-HRV measured in 2007 was associated with higher self-reported sleep complaints 4 years later after controlling for covariates (rab,c = -0.096, b = -0.108, 95% CI, -0.298 to 0.081, ΔR2 = 0.009, p = .050). These data are the first to show that lower sleep HF-HRV predicted worse sleep 4 years later, highlighting the importance of vagus nerve functioning in adaptability and health.

Discrete Event Systems Theory for Fast Stochastic Simulation via Tree Expansion

Discrete Event Systems Theory for Fast Stochastic Simulation via Tree Expansion

The Childhood Decade in Russia: legal securing of health care applying information technologies

In conditions of impetuous development of national digital technologies and necessity for their legal regulation, significance and possibilities of their implementation in protecting health of rising generation during childhood decade in Russia are demonstrated. The conceptual apparatus related to discussed problem is considered. The emphasis is made on preventive aspect of intersectoral state protection of health care and rights of children and adolescents aged 0-17 years. The scale of problem is especially impressive considering number of contingent over 30 million people and interests of about 24 million of family members with children. The purpose of the study is applying analytical method, content analysis and SWOT analysis to monitor main problems of intersectoral health care and rights of rising generation in Russia in conditions of implementation of information technologies; to evaluate application of legislation in force in context of their role and effect during the Decade of Children (2018-2027); to demonstrate significance of the institution of the Commissioner for rights of child under the President of the Russian Federation.

Privacy preserving image registration

Image registration is a key task in medical imaging applications, allowing to represent medical images in a common spatial reference frame. Current approaches to image registration are generally based on the assumption that the content of the images is usually accessible in clear form, from which the spatial transformation is subsequently estimated. This common assumption may not be met in practical applications, since the sensitive nature of medical images may ultimately require their analysis under privacy constraints, preventing to openly share the image content. In this work, we formulate the problem of image registration under a privacy preserving regime, where images are assumed to be confidential and cannot be disclosed in clear. We derive our privacy preserving image registration framework by extending classical registration paradigms to account for advanced cryptographic tools, such as secure multi-party computation and homomorphic encryption, that enable the execution of operations without leaking the underlying data. To overcome the problem of performance and scalability of cryptographic tools in high dimensions, we propose several techniques to optimize the image registration operations by using gradient approximations, and by revisiting the use of homomorphic encryption trough packing, to allow the efficient encryption and multiplication of large matrices. We focus on registration methods of increasing complexity, including rigid, affine, and non-linear registration based on cubic splines or diffeomorphisms parameterized by time-varying velocity fields. In all these settings, we demonstrate how the registration problem can be naturally adapted for accounting to privacy-preserving operations, and illustrate the effectiveness of PPIR on a variety of registration tasks.

RELIABILITY ASSESSMENT OF STEEL PORTAL FRAMES USING GAN FOR GENERATING SYNTHETIC DATA SAMPLE

AbstractThis paper addresses the challenge of reliability estimation in structural engineering, where determining failure probabilities is often uncertain due to insufficient characterization of randomness and scale of design problems. Current approaches, like EN 1990, lack sufficient data for research and simulations, particularly for low failure probabilities (around 10‐4), making Monte Carlo simulations less accurate. The paper introduces Generative Adversarial Networks (GANs) as a solution to generate synthetic data to supplement existing examples. The study applies GANs to assess the reliability of steel portal‐framed industrial buildings and evaluate the safety of this structural solution according to Eurocodes.

Interactive crowdsourcing to fact-check politicians.

The discourse of political leaders often contains false information that can misguide the public. Fact-checking agencies around the world try to reduce the negative influence of politicians by verifying their words. However, these agencies face a problem of scalability and require innovative solutions to deal with their growing amount of work. While the previous studies have shown that crowdsourcing is a promising approach to fact-check news in a scalable manner, it remains unclear whether crowdsourced judgements are useful to verify the speech of politicians. This article fills that gap by studying the effect of social influence on the accuracy of collective judgements about the veracity of political speech. In this work, we performed two experiments (Study 1: N = 180; Study 2: N = 240) where participants judged the veracity of 20 politically balanced phrases. Then, they were exposed to social information from politically homogeneous or heterogeneous participants. Finally, they provided revised individual judgements. We found that only heterogeneous social influence increased the accuracy of participants compared to a control condition. Overall, our results uncover the effect of social influence on the accuracy of collective judgements about the veracity of political speech and show how interactive crowdsourcing strategies can help fact-checking agencies. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

A novel multi-scale CNN and Bi-LSTM arbitration dense network model for low-rate DDoS attack detection

Low-rate distributed denial of service attacks, as known as LDDoS attacks, pose the notorious security risks in cloud computing network. They overload the cloud servers and degrade network service quality with the stealthy strategy. Furthermore, this kind of small ratio and pulse-like abnormal traffic leads to a serious data scale problem. As a result, the existing models for detecting minority and adversary LDDoS attacks are insufficient in both detection accuracy and time consumption. This paper proposes a novel multi-scale Convolutional Neural Networks (CNN) and bidirectional Long-short Term Memory (bi-LSTM) arbitration dense network model (called MSCBL-ADN) for learning and detecting LDDoS attack behaviors under the condition of limited dataset and time consumption. The MSCBL-ADN incorporates CNN for preliminary spatial feature extraction and embedding-based bi-LSTM for time relationship extraction. And then, it employs arbitration network to re-weigh feature importance for higher accuracy. At last, it uses 2-block dense connection network to perform final classification. The experimental results conducted on popular ISCX-2016-SlowDos dataset have demonstrated that the proposed MSCBL-ADN model has a significant improvement with high detection accuracy and superior time performance over the state-of-the-art models.

Psychometric properties and normative data of the Latvian and Russian language versions of the strengths and difficulties questionnaire (SDQ) in the Latvian general adolescent population

Objectives Mental health screening instruments are essential in population health research and clinical practice. The strengths and difficulties questionnaire (SDQ) self-report version has been widely used across the globe to screen for mental health problems in adolescent populations. This study aimed to explore the psychometric properties of the Latvian and Russian language versions of the SDQ in a representative sample of a general population of Latvian adolescents and establish the population-based normative scores. Materials and methods This analysis was based on data from the international Health Behaviour in School-aged Children study year 2017/2018 Latvian database. The sample comprised 2683 Latvian and 1321 Russian-speaking 11-, 13-, and 15-year-old adolescents. Results Significant language-based variance was observed in the mean scores of the externalising subscales of the SDQ, with Latvian-speaking adolescents reporting higher levels of conduct problems and hyperactivity. The reliability was satisfactory (ordinal alpha >0.7) only for the prosocial behaviour, emotional, internalising problems, and total difficulties subscales, while conduct, hyperactivity, peer, and externalising problems scales demonstrated lower internal consistency (ordinal alpha 0.5–0.7). Confirmatory factor analysis (CFA) did not support the hypothesised narrow-band five-factor or broad-band three-factor structure of the SDQ. The normative banding scores were calculated for both language variants. Conclusions This study suggests that the self-report SDQ must be used cautiously in Latvian adolescents because some SDQ subscales lack reliability. When used in population research, the internalising subscales seem more reliable than the externalising ones. More research is needed regarding the reliability of the SDQ in clinical adolescent populations.

Spatial Accessibility to Primary Care in Metropolitan France: Results Using the SCALE Spatial Accessibility Index for All Regions.

Accessibility to care is a major public health issue. Various tools to assess it are available, but they do not solve the problem of scale. Moreover, accessibility is a multidimensional concept that is not taken into account with current tools. The SCALE index aims to overcome these two limitations by proposing a synthetic measure on a more precise scale than the administrative unit or the sub-municipal scale. Under the assumption of access to care facilities for all and access to the nearest facilities, the potential accessibility distance was calculated for each couple (residential area, accessible facilities). This was defined as the average distance by road that the population has to travel to access care. To take the availability of resources into account, these distances were weighted by the theoretical pressure on the facilities. The SCALE index was then calculated using a linear combination of the distances of potential accessibility to care facilities It highlights differences in accessibility at the national and regional scale. Using this index, it was possible to provide maps for all French regions and the major cities in a story-map. The major conurbation around Paris and the main urban centers has high accessibility. Low accessibility forms a "Y" shape. In conclusion, the SCALE index measures accessibility at the scale of a small geographic unit taking the proximity and the availability of health professionals into account. It is also possible to take into account the diversity of accessibility in a given territory.

Asymmetric Learning for Graph Neural Network based Link Prediction

Link prediction is a fundamental problem in many graph-based applications, such as protein-protein interaction prediction. Recently, graph neural network (GNN) has been widely used for link prediction. However, existing GNN-based link prediction (GNN-LP) methods suffer from scalability problem during training for large-scale graphs, which has received little attention from researchers. In this paper, we first analyze the computation complexity of existing GNN-LP methods, revealing that one reason for the scalability problem stems from their symmetric learning strategy in applying the same class of GNN models to learn representation for both head nodes and tail nodes. We then propose a novel method, called a sym m etric l earning (AML), for GNN-LP. More specifically, AML applies a GNN model to learn head node representation while applying a multi-layer perceptron (MLP) model to learn tail node representation. To the best of our knowledge, AML is the first GNN-LP method to adopt an asymmetric learning strategy for node representation learning. Furthermore, we design a novel model architecture and apply a row-wise mini-batch sampling strategy to ensure promising model accuracy and training efficiency for AML. Experiments on three real large-scale datasets show that AML is 1.7×∼7.3× faster in training than baselines with a symmetric learning strategy while having almost no accuracy loss.