Search

Graph embedding has gained significant popularity due to its ability to represent large-scale graph data by mapping nodes to a low-dimensional space. However, most of the existing research in this field has focused on transductive learning, where fixed node embeddings are generated by training the entire graph. This approach is not well-suited for temporal graphs that undergo continuous changes with the addition of new nodes and interactions. To address this limitation, we propose an inductive temporal graph embedding method called MIAN (Multi-angle Information Aggregation Network). The key focus of MIAN is to design an aggregation function that combines multi-angle information for generating node embeddings. Specifically, we divide the information into different angles, including neighborhood, temporal, and environment. Each angle of information is modeled and mined independently, and then fed into an improved gated recuttent unit (GRU) module to effectively combine them. To assess the performance of MIAN, we conduct extensive experiments on various real-world datasets and compare its results with several state-of-the-art baseline methods across diverse tasks. The experimental findings demonstrate that MIAN outperforms these methods. [ABSTRACT FROM AUTHOR]

Benefits of self-testing for learning have been consistently shown for simple materials such as word lists learned by rote memorization. Considerably less evidence for such benefits exists for complex, more educationally relevant materials and its application to new situations. The present study explores the mechanisms underlying this transfer. To this end, a typical retrieval-practice-effect paradigm was applied to foster the learning of an artificial language. Participants either repeatedly studied grammatically correct exemplar sentences of the artificial language or engaged in a cloze test as the interim test after learning. To assess far transfer, participants in both groups of restudy and retrieval practice engaged in a grammaticality judgment test after a delay of 5 min and 1 week. In addition, participants in both groups completed a final memory test (i.e., a cloze test identical to the initial test) 1 week after learning. In addition to a long-term memory benefit of retrieval practice, results revealed also a retrieval-practice benefit in the far-transfer test after the 1-week delay. The findings further support the view that far transfer is supported by learning the underlying grammatical rules as opposed to memorizing the material. Thus, retrieval practice is also effective for fostering learning of complex materials and, even more importantly, for promoting transfer of learning—a crucial goal in modern educational practices. [ABSTRACT FROM AUTHOR]

Network science has witnessed a surge in popularity, driven by the transformative power of node representation learning for diverse applications like social network analysis and biological modeling. While shallow embedding algorithms excel at capturing network structure, they face a critical limitation—failing to generalize to unseen nodes. This paper addresses this challenge by introducing Inductive Shallow Node Embedding—as a main contribution—pioneering a novel approach that extends shallow embeddings to the realm of inductive learning. It has a novel encoder architecture that captures the local neighborhood structure of each node, enabling effective generalization to unseen nodes. In the generalization, robustness is essential to avoid degradation of performance arising from noise in the dataset. It has been theoretically proven that the covariance of the additive noise term in the proposed model is inversely proportional to the cardinality of a node's neighbors. Another contribution is a mathematical lower bound to quantify the robustness of node embeddings, confirming its advantage over traditional shallow embedding methods, particularly in the presence of parameter noise. The proposed method demonstrably excels in dynamic networks, consistently achieving over 90% performance on previously unseen nodes compared to nodes encountered during training on various benchmarks. The empirical evaluation concludes that our method outperforms competing methods on the vast majority of datasets in both transductive and inductive tasks. [ABSTRACT FROM AUTHOR]

Abstract Benefits of self-testing for learning have been consistently shown for simple materials such as word lists learned by rote memorization. Considerably less evidence for such benefits exists for complex, more educationally relevant materials and its application to new situations. The present study explores the mechanisms underlying this transfer. To this end, a typical retrieval-practice-effect paradigm was applied to foster the learning of an artificial language. Participants either repeatedly studied grammatically correct exemplar sentences of the artificial language or engaged in a cloze test as the interim test after learning. To assess far transfer, participants in both groups of restudy and retrieval practice engaged in a grammaticality judgment test after a delay of 5 min and 1 week. In addition, participants in both groups completed a final memory test (i.e., a cloze test identical to the initial test) 1 week after learning. In addition to a long-term memory benefit of retrieval practice, results revealed also a retrieval-practice benefit in the far-transfer test after the 1-week delay. The findings further support the view that far transfer is supported by learning the underlying grammatical rules as opposed to memorizing the material. Thus, retrieval practice is also effective for fostering learning of complex materials and, even more importantly, for promoting transfer of learning—a crucial goal in modern educational practices.

Abstract Network science has witnessed a surge in popularity, driven by the transformative power of node representation learning for diverse applications like social network analysis and biological modeling. While shallow embedding algorithms excel at capturing network structure, they face a critical limitation—failing to generalize to unseen nodes. This paper addresses this challenge by introducing Inductive Shallow Node Embedding—as a main contribution—pioneering a novel approach that extends shallow embeddings to the realm of inductive learning. It has a novel encoder architecture that captures the local neighborhood structure of each node, enabling effective generalization to unseen nodes. In the generalization, robustness is essential to avoid degradation of performance arising from noise in the dataset. It has been theoretically proven that the covariance of the additive noise term in the proposed model is inversely proportional to the cardinality of a node’s neighbors. Another contribution is a mathematical lower bound to quantify the robustness of node embeddings, confirming its advantage over traditional shallow embedding methods, particularly in the presence of parameter noise. The proposed method demonstrably excels in dynamic networks, consistently achieving over 90% performance on previously unseen nodes compared to nodes encountered during training on various benchmarks. The empirical evaluation concludes that our method outperforms competing methods on the vast majority of datasets in both transductive and inductive tasks.

The interaction between T-cell receptors (TCRs) and peptides (epitopes) presented by major histocompatibility complex molecules (MHC) is fundamental to the immune response. Accurate prediction of TCR–epitope interactions is crucial for advancing the understanding of various diseases and their prevention and treatment. Existing methods primarily rely on sequence-based approaches, overlooking the inherent topology structure of TCR–epitope interaction networks. In this study, we present |$GTE$|⁠ , a novel heterogeneous Graph neural network model based on inductive learning to capture the topological structure between TCRs and Epitopes. Furthermore, we address the challenge of constructing negative samples within the graph by proposing a dynamic edge update strategy, enhancing model learning with the nonbinding TCR–epitope pairs. Additionally, to overcome data imbalance, we adapt the Deep AUC Maximization strategy to the graph domain. Extensive experiments are conducted on four public datasets to demonstrate the superiority of exploring underlying topological structures in predicting TCR–epitope interactions, illustrating the benefits of delving into complex molecular networks. The implementation code and data are available at https://github.com/uta-smile/GTE. [ABSTRACT FROM AUTHOR]

Graph embedding has gained significant popularity due to its ability to represent large-scale graph data by mapping nodes to a low-dimensional space. However, most of the existing research in this field has focused on transductive learning, where fixed node embeddings are generated by training the entire graph. This approach is not well-suited for temporal graphs that undergo continuous changes with the addition of new nodes and interactions. To address this limitation, we propose an inductive temporal graph embedding method called MIAN (Multi-angle Information Aggregation Network). The key focus of MIAN is to design an aggregation function that combines multi-angle information for generating node embeddings. Specifically, we divide the information into different angles, including neighborhood, temporal, and environment. Each angle of information is modeled and mined independently, and then fed into an improved gated recuttent unit (GRU) module to effectively combine them. To assess the performance of MIAN, we conduct extensive experiments on various real-world datasets and compare its results with several state-of-the-art baseline methods across diverse tasks. The experimental findings demonstrate that MIAN outperforms these methods.

Knowledge graph completion can make knowledge graph more complete. However, traditional knowledge graph completion methods assume that all test entities and relations appear in the training process. Due to the evolving nature of real world KG, once unseen entities or relations appear, the knowledge graph needs to be retrained. Inductive learning for knowledge graph completion aims to complete triples containing unseen entities or unseen relations without training the knowledge graph from scratch, so it has received much attention in recent years. Firstly, starting from the basic concept of knowledge graph, this paper divides knowledge graph completion into two categories: transductive and inductive. Secondly, from the theoretical perspective of inductive knowledge graph completion, it is divided into two categories: semi-inductive and fully-inductive, and the models are summarized from this perspective. Then, from the technical perspective of inductive knowledge graph completion, it is divided into two categories: based on structural information and based on additional information. The methods based on structural information are subdivided into three categories: based on inductive embedding, based on logical rules and based on meta learning, and the methods based on additional information are subdivided into two categories: based on text information and other information. The current methods are further subdivided, analyzed and compared. Finally, it forecasts the main research directions in the future.

Copyright of Journal of Frontiers of Computer Science & Technology is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Demands for more accurate machine learning models have given rise to rethinking current modeling approaches that were deemed unsuitable, primarily due to their computational complexity and the lack of availability and accessibility to representative data. In Industry 4.0, rapid advancements in Digital Twin (DT) technologies and the pervasiveness of cost-effective sensor technologies have pushed the incorporation of artificial intelligence, particularly data-driven machine learning models, for use in smart manufacturing. However, the persistent issue with such models is their high sensitivity to the training data and the lack of interpretability in the outcomes, at times generating unrealistic results. The incorporation of knowledge into the machine learning pipeline has been earmarked as the most promising approach to address such issues. This paper aims to answer this call through a Knowledge-embedded Machine Learning (KML) framework for smart manufacturing, which embeds knowledge from experience and, or physics information into the machine learning pipeline, thus making the outcomes from these models more representative of real applications. The merits of KML were then presented through comparative studies showing its capability to outperform knowledge-based and data-driven models. This promising outcome led to the development of frameworks that can potentially incorporate KML for smart manufacturing applications such as Prognostics and Health Management (PHM) and DT, further supporting the usefulness of the proposed KML framework. [ABSTRACT FROM AUTHOR]

Research has demonstrated that L2 English learners' acquisition of grammar can be facilitated by Data-Driven Learning (DDL), which clearly promotes inductive learning and learner-centeredness (e.g. Boontam & Phoochaoroensil, 2018). This present study focuses on the impacts of DDL on Thai secondary and tertiary educators' introduction of English conditionals. The participants were taught using corpus-based DDL materials representing authentic use of conditionals for six weeks. It was shown in the interview results that DDL is useful and enhances learners' self-discovery of L2 grammar rules regarding conditionals. Not only does DDL enable learners to notice the three classic if-conditional types but also it allows them to explore alternative, i.e. mixed, types of conditionals, which are often overlooked by textbooks. Some concerns about the use of DDL in a broader context of ELT deal with school budgets and students' need of technical help. [ABSTRACT FROM AUTHOR]

Cyber-attacks pose increasing challenges in precisely detecting intrusions, risking data confidentiality, integrity, and availability. This review paper presents recent IDS taxonomy, a comprehensive review of intrusion detection techniques, and commonly used datasets for evaluation. It discusses evasion techniques employed by attackers and the challenges in combating them to enhance network security. Researchers strive to improve IDS by accurately detecting intruders, reducing false positives, and identifying new threats. Machine learning (ML) and deep learning (DL) techniques are adopted in IDS systems, showing potential in efficiently detecting intruders across networks. The paper explores the latest trends and advancements in ML and DL-based network intrusion detection systems (NIDS), including methodology, evaluation metrics, and dataset selection. It emphasizes research obstacles and proposes a future research model to address weaknesses in the methodologies. The decision tree, known for its speed and user-friendliness, is proposed as a model for detecting result anomalies, combining findings from a comparative survey. This research aims to provide insights into building an effective decision tree-based detection framework.

La Elasticidad y Resistencia de Materiales (ERM) es una disciplina impartida en muchos grados de ingeniería. Esta asignatura necesita que los alumnos presenten una actitud participativa durante el proceso de aprendizaje, debido a su complejidad e importancia a lo largo del grado. Sin embargo, los estudiantes la encuentran muy complicada y abstracta. Por tanto, la principal dificultad que encuentra el profesor a la hora de enseñar es mantener el interés y motivación de los alumnos durante el proceso de aprendizaje. Para ayudar en esta tarea, los profesores deben ayudarse de nuevas metodologías como el aprendizaje basado en problemas y la clase inversa, donde se facilita la información a los estudiantes previamente a la clase presencial. En este trabajo, se presenta la aplicación de un aprendizaje basado en problemas al laboratorio de ERM en el grado de Ingeniería Mecánica. Además, para aumentar el entusiasmo y la motivación de los estudiantes, estas sesiones de laboratorio incluyeron una tecnología de fabricación innovadora, impresión 3D, y el empleo de correlación de imágenes digitales (DIC) para medida de desplazamientos y deformaciones. Antes de cada sesión práctica, se animaba a los alumnos a visualizar un vídeo online con los aspectos fundamentales de la práctica. Para evaluar el éxito de esta metodología, después de terminar las sesiones de laboratorio, los estudiantes respondieron una encuesta cuantitativa no formal. Los resultados mostraron que el aprendizaje basado en problemas propuesto tenía la capacidad de ayudar a integrar los conocimientos y mejorar la adquisición de las competencias incluidas en la guía docente de la asignatura. Aunque estos resultados son alentadores, todavía hay partes de la actividad de laboratorio que deben mejorarse para que la actividad consuma menos tiempo y facilitar las partes más complejas para los estudiantes. /// Elasticity and Strength of Materials is an elementary discipline across many engineering degrees. Students perceive it as very difficult and abstract and students’ active participation in the learning process is needed, due to its complexity and importance during the whole degree years. Hence, the main difficulty in teaching them is maintaining students' motivation and keeping their participation in the learning process. This is the reason to use new tools, such as problem-based learning (PBL) and flipped classroom, to help students understand and study the subject. In this work, we present the application of a problem-based learning to the lab in the Mechanical engineering degree, in of one the most time-consuming and difficult subjects of its program. Moreover, to enhance students’ enthusiasm and motivation, these sessions included an innovative manufacturing technology, 3D printing, and digital image correlation (DIC). Before each practical session, the students are encouraged to watch an online video with the fundamental aspects. In order to assess the success of this laboratory course, after finishing the lab sessions, the students answered a non-formal quantitative survey. The results showed that the proposed project-based learning had the ability to help integrating the knowledge and improve the skills included in the main competences. Although these results are encouraging, there are still parts of the lab activity that should be improve in order to make the activity less time consuming and the most difficult part being easier for the students.

The performance of optimization algorithms, and consequently of AI/machine learning solutions, is strongly influenced by the setting of their hyperparameters. Over the last decades, a rich literature has developed proposing methods to automatically determine the parameter setting for a problem of interest, aiming at either robust or instance-specific settings. Robust setting optimization is already a mature area of research, while instance-level setting is still in its infancy, with contributions mainly dealing with algorithm selection. The work reported in this paper belongs to the latter category, exploiting the learning and generalization capabilities of artificial neural networks to adapt a general setting generated by state-of-the-art automatic configurators. Our approach differs significantly from analogous ones in the literature, both because we rely on neural systems to suggest the settings, and because we propose a novel learning scheme in which different outputs are proposed for each input, in order to support generalization from examples. The approach was validated on two different algorithms that optimized instances of two different problems. We used an algorithm that is very sensitive to parameter settings, applied to generalized assignment problem instances, and a robust tabu search that is purportedly little sensitive to its settings, applied to quadratic assignment problem instances. The computational results in both cases attest to the effectiveness of the approach, especially when applied to instances that are structurally very different from those previously encountered. [ABSTRACT FROM AUTHOR]

Research on sequence effects on learning visual categories has shown that interleaving (i.e., studying the categories in a mixed manner) facilitates category induction as compared to blocking (i.e., studying the categories one by one), but learners are unaware of the interleaving effect and prefer blocking. However, little attention has been paid to sequence effects in perceptual learning across further sensory modalities. The present (preregistered) research addresses this shortcoming by using auditory (birdcalls), olfactory (tealeaves), gustatory (ingredient mixtures), and tactile (stones) stimuli across four analog experiments. The number of participants per experiment was determined based on a medium effect size of interleaving. Participants studied six categories (with six exemplars, respectively) either interleaved or blocked. No single experiment showed a significant effect of interleaving. We ran a comprehensive meta‐analysis based on the data from all experiments, which revealed a significant small effect of interleaving, demonstrating its applicability to perceptual learning across all sensory modalities. Learners in the interleaved condition underestimated their classification performance. Overall, learners did not rate interleaving as less effective than blocking, which is at odds with previous studies that consistently demonstrated a metacognitive preference of blocking. Our findings suggest that learners rely less on conventional beliefs about the effectivity of study sequence when dealing with unfamiliar (blindfolded) perceptual learning tasks. [ABSTRACT FROM AUTHOR]

This article introduces ILA4: A new algorithm designed to handle datasets with missing values. ILA4 is inspired by a series of ILA algorithms which also handle missing data with further enhancements. ILA4 is applied to datasets with varying completeness and also compared to other, known approaches for handling datasets with missing values. In the majority of cases, ILA4 produced favorable performance that is on a par with many established approaches for treating missing values including algorithms that are based on the Most Common Value (MCV), the Most Common Value Restricted to a Concept (MCVRC), and those that utilize the Delete strategy. ILA4 was also compared with three known algorithms namely: Logistic Regression, Naïve Bayes, and Random Forest; the accuracy obtained by ILA4 is comparable or better than the best results obtained from these three algorithms.

Collaborative learning entails the involvement and the cooperation of a group of persons with the purpose of learning. Collaborative learning scripts aim to orchestrate the complex interaction among group members while Computer Supported Collaborative Learning scripts (CSCL scripts) is the research field in which IT techniques are involved in the management of the aspects of such an interaction. This article presents assessment-related aspects of an existing CSCL script authoring and deployment platform called COSTLyP. Assessment, nowadays, is considered as a vital constituent of CSCL scripts since it may affect some of their necessary components and mechanisms. The outcome of the implementation of an assessment plan may determine what should be the next step in a collaboration activity or what actions should be undertaken to bridge the gap between the expected results and the achieved level of knowledge or expertise. At the same time, assessment can also verify the regulation level that is required within each group; consequently, these scripts should be flexibly designed in order to adapt their evolution to the real needs of the participants. [ABSTRACT FROM AUTHOR]

Elasticity and Strength of Materials is an elementary discipline across many engineering degrees. Students perceive it as very difficult and abstract and students' active participation in the learning process is needed, due to its complexity and importance during the whole degree years. Hence, the main difficulty in teaching them is maintaining students' motivation and keeping their participation in the learning process. This is the reason to use new tools, such as problem-based learning (PBL) and flipped classroom, to help students understand and study the subject. In this work, we present the application of a problem-based learning to the lab in the Mechanical engineering degree, in of one the most time-consuming and difficult subjects of its program. Moreover, to enhance students' enthusiasm and motivation, these sessions included an innovative manufacturing technology, 3D printing, and digital image correlation (DIC). Before each practical session, the students are encouraged to watch an online video with the fundamental aspects. In order to assess the success of this laboratory course, after finishing the lab sessions, the students answered a non-formal quantitative survey. The results showed that the proposed project-based learning had the ability to help integrating the knowledge and improve the skills included in the main competences. Although these results are encouraging, there are still parts of the lab activity that should be improve in order to make the activity less time consuming and the most difficult part being easier for the students. [ABSTRACT FROM AUTHOR]

Unsupervised domain adaptation (UDA) is to learn classification models that make predictions for unlabeled data on a target domain, given labeled data on a source domain whose distribution diverges from the target one. Mainstream UDA methods strive to learn domain-aligned features such that classifiers trained on the source features can be readily applied to the target ones. Although impressive results have been achieved, these methods have a potential risk of damaging the intrinsic data structures of target discrimination, raising an issue of generalization particularly for UDA tasks in an inductive setting. To address this issue, we are motivated by a UDA assumption of structural similarity across domains, and propose to directly uncover the intrinsic target discrimination via constrained clustering, where we constrain the clustering solutions using structural source regularization that hinges on the very same assumption. Technically, we propose a hybrid model of Structurally Regularized Deep Clustering, which integrates the regularized discriminative clustering of target data with a generative one, and we thus term our method as H-SRDC. Our hybrid model is based on a deep clustering framework that minimizes the Kullback-Leibler divergence between the distribution of network prediction and an auxiliary one, where we impose structural regularization by learning domain-shared classifier and cluster centroids. By enriching the structural similarity assumption, we are able to extend H-SRDC for a pixel-level UDA task of semantic segmentation. We conduct extensive experiments on seven UDA benchmarks of image classification and semantic segmentation. With no explicit feature alignment, our proposed H-SRDC outperforms all the existing methods under both the inductive and transductive settings. We make our implementation codes publicly available at https://github.com/huitangtang/H-SRDC. [ABSTRACT FROM AUTHOR]

This article introduces ILA4: A new algorithm designed to handle datasets with missing values. ILA4 is inspired by a series of ILA algorithms which also handle missing data with further enhancements. ILA4 is applied to datasets with varying completeness and also compared to other, known approaches for handling datasets with missing values. In the majority of cases, ILA4 produced favorable performance that is on a par with many established approaches for treating missing values including algorithms that are based on the Most Common Value (MCV), the Most Common Value Restricted to a Concept (MCVRC), and those that utilize the Delete strategy. ILA4 was also compared with three known algorithms namely: Logistic Regression, Naïve Bayes, and Random Forest; the accuracy obtained by ILA4 is comparable or better than the best results obtained from these three algorithms. [ABSTRACT FROM AUTHOR]

The performance of optimization algorithms, and consequently of AI/machine learning solutions, is strongly influenced by the setting of their hyperparameters. Over the last decades, a rich literature has developed proposing methods to automatically determine the parameter setting for a problem of interest, aiming at either robust or instance-specific settings. Robust setting optimization is already a mature area of research, while instance-level setting is still in its infancy, with contributions mainly dealing with algorithm selection. The work reported in this paper belongs to the latter category, exploiting the learning and generalization capabilities of artificial neural networks to adapt a general setting generated by state-of-the-art automatic configurators. Our approach differs significantly from analogous ones in the literature, both because we rely on neural systems to suggest the settings, and because we propose a novel learning scheme in which different outputs are proposed for each input, in order to support generalization from examples. The approach was validated on two different algorithms that optimized instances of two different problems. We used an algorithm that is very sensitive to parameter settings, applied to generalized assignment problem instances, and a robust tabu search that is purportedly little sensitive to its settings, applied to quadratic assignment problem instances. The computational results in both cases attest to the effectiveness of the approach, especially when applied to instances that are structurally very different from those previously encountered.

Soft errors are increasing in computer systems due to shrinking feature sizes. Soft errors can induce incorrect outputs, also called silent data corruption (SDC), which raises no warnings in the system and hence is difficult to detect. To prevent SDC effectively, protection techniques require a fine-grained profiling of SDC-prone instructions, which is often obtained by applying machine learning models. However, these models rely on handcrafted features, and lack the ability to reason about SDC propagation, which leads to an inferior SDC prediction performance. We propose a novel Graph Attention neTwork to Predict SDC-prone instructions (GATPS). The GATPS representation is a heterogeneous graph with different types of edges to represent various instruction relations. By stacking layers in which nodes are able to attend over their neighborhoods' features, GATPS automatically captures the structural features that contribute to SDC propagation. The attention mechanism is applied to compute the importance values to the neighboring nodes, which quantifies the fault effect on the neighboring nodes. Moreover, the inductive model of GATPS can be applied to unseen programs without retraining, and it requires no fault injection information of the target program. Experiments revealed GATPS achieved a 34% higher F1 score compared to the baseline method and a 40-fold speedup compared to the fault injection approach. [ABSTRACT FROM AUTHOR]

Many real–world domains involve information naturally represented by graphs, where nodes denote basic patterns while edges stand for relationships among them. The graph neural network (GNN) is a machine learning model capable of directly managing graph–structured data. In the original framework, GNNs are inductively trained, adapting their parameters based on a supervised learning environment. However, GNNs can also take advantage of transductive learning, thanks to the natural way they make information flow and spread across the graph, using relationships among patterns. In this paper, we propose a mixed inductive–transductive GNN model, study its properties and introduce an experimental strategy that allows us to understand and distinguish the role of inductive and transductive learning. The preliminary experimental results show interesting properties for the mixed model, highlighting how the peculiarities of the problems and the data can impact on the two learning strategies. [ABSTRACT FROM AUTHOR]

Copyright of InTRAlinea: Online Translation Journal is the property of Department Interpreting & Translation (DIT) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Discovering gene regulatory relationships and reconstructing gene regulatory networks (GRN) based on gene expression data is a classical, long-standing computational challenge in bioinformatics. Computationally inferring a possible regulatory relationship between two genes can be formulated as a link prediction problem between two nodes in a graph. Graph neural network (GNN) provides an opportunity to construct GRN by integrating topological neighbor propagation through the whole gene network. We propose an end-to-end gene regulatory graph neural network (GRGNN) approach to reconstruct GRNs from scratch utilizing the gene expression data, in both a supervised and a semi-supervised framework. To get better inductive generalization capability, GRN inference is formulated as a graph classification problem, to distinguish whether a subgraph centered at two nodes contains the link between the two nodes. A linked pair between a transcription factor (TF) and a target gene, and their neighbors are labeled as a positive subgraph, while an unlinked TF and target gene pair and their neighbors are labeled as a negative subgraph. A GNN model is constructed with node features from both explicit gene expression and graph embedding. We demonstrate a noisy starting graph structure built from partial information, such as Pearson’s correlation coefficient and mutual information can help guide the GRN inference through an appropriate ensemble technique. Furthermore, a semi-supervised scheme is implemented to increase the quality of the classifier. When compared with established methods, GRGNN achieved state-of-the-art performance on the DREAM5 GRN inference benchmarks. GRGNN is publicly available at https://github.com/juexinwang/GRGNN.

Applying machine learning techniques to spatial-temporal data poses the question that how the recorded location and time for training samples should contribute to the training and testing process. The prior knowledge of how spatial-temporal phenomena are autocorrelated cannot be properly captured by machine learning techniques, which either ignore location and time altogether or consider them as input features. Not to mention that the latter approach leads to slightly increased sparseness of data in the feature space and more free parameters in the predictor; thus, demanding for larger training datasets. We use the prior knowledge about the spatial-temporal autocorrelation to determine how relevant each training sample would be, given its spatial and temporal distances to the irresponsive (unlabeled) sample. Weighted machine learning techniques use this prior knowledge by taking the relevance of training samples with regard to the irresponsive sample into account as training samples' weights. The proposed approach overcomes the aforementioned issues by enriching the training process with the prior knowledge about spatial-temporal autocorrelation. Because the spatial-temporal weight of training samples depends on the irresponsive sample's location and time, the machine needs to be trained separately for each irresponsive sample. However, we show that in practice using only a small subset of training samples with largest spatial-temporal weights not only mitigates the training time but also results in the best accuracy in most cases.

Collaborative learning entails the involvement and the cooperation of a group of persons with the purpose of learning. Collaborative learning scripts aim to orchestrate the complex interaction among group members while Computer Supported Collaborative Learning scripts (CSCL scripts) is the research field in which IT techniques are involved in the management of the aspects of such an interaction. This article presents assessment-related aspects of an existing CSCL script authoring and deployment platform called COSTLyP. Assessment, nowadays, is considered as a vital constituent of CSCL scripts since it may affect some of their necessary components and mechanisms. The outcome of the implementation of an assessment plan may determine what should be the next step in a collaboration activity or what actions should be undertaken to bridge the gap between the expected results and the achieved level of knowledge or expertise. At the same time, assessment can also verify the regulation level that is required within each group; consequently, these scripts should be flexibly designed in order to adapt their evolution to the real needs of the participants.

Graph carries out a key role in graph-based semi-supervised label propagation, as it clarifies the structure of the data manifold. The performance of label propagation methods depends on the adopted graph and can be enhanced by merging different graphs that are obtained from multiple sources of information. While there exist algorithms that perform graph fusion they have several weaknesses. Most of these algorithms define graph fusion and label propagation as two separate tasks. Moreover, when the number of data expands, these strategies are not well-suited due to the use of transductive learning in the label propagation phase which makes the label prediction for unseen samples difficult. Furthermore, very few algorithms extract the information contained in the label space. Additionally, most of the graph fusion techniques adopt equal or static weights for different views, which is not the best choice as distinctive features (hence different graphs) contain various information. To overcome these shortcomings, we propose an Auto-weighted Multi-view Semi-Supervised Learning method (AMSSL), which is based on an inductive learning algorithm (i.e., Flexible Manifold Embedding) and profited a projection matrix for predicting the labels of out-of-sample data. The proposed AMSSL method represents a unified framework that dynamically fuses various information obtained from different features and also from the label space and adaptively designates appropriate weights according to the usefulness of each view. Our experimental results on seven small and large image datasets demonstrate the superiority of the proposed method compared to the use of one single feature and other state-of-the-art graph fusion methods. • We propose an algorithm which jointly fuses the graphs and estimates the labels. • Suitable for large scale image classification. • Adaptively assigns appropriate weights to different views. • Explicitly uses the information in the label space. • Results on six databases prove the outperformance of the proposed method. [ABSTRACT FROM AUTHOR]

Contribution: A novel project-based learning (PBL) approach applicable to diverse design-oriented engineering courses. This article reports the application results during six academic years (2012/13 to 2017/18) for electronic instrumentation (EI) course for third-year Industrial Electronic and Automation Engineering majors. Background: Engineering education should prepare students to specify, design, and validate systems interacting with stakeholders in complex professional scenarios. Intended Outcomes: The general aim of the course was to train students in the specification, design, and validation of real-world instrumentation systems, promoting the development of expertise and deep understanding of fundamental concepts and techniques. Application Design: The students, working in teams, were tasked with the specification, design, and development of different instrumentation projects. They had to exchange information about their projects with other teams and discuss design decisions, thus creating a multiple-case experience. In addition, each team acted as a provider of a different team and as a client of another team, in a circular client–provider relationship. This student-centered learning process was supported and complemented with guided teaching of fundamental concepts, principles, and techniques and with continuous evaluation, complemented with rigorous individual exams. The project activities, lectures, and problem-solving classes were coordinated to optimize the process of developing the students’ understanding and expertise. Findings: Based on the academic results and the feedback from the students obtained along the years through two different anonymous surveys, it is concluded that most of the students achieved the expected learning outcomes, i.e., learning to design EI systems and dealing with stakeholders. Analyzed results proved that students were very satisfied with the methodology. [ABSTRACT FROM AUTHOR]

Industrial design education is focused on teaching a combination of various interdisciplinary competencies. One of these projected learning outcomes is to be able to design mechanisms in order to fulfil certain mechanical constraints in products. Studies show that theoretical knowledge supported by practice helps to teach industrial design students the mechanisms. In the current situation in Turkey, practice-based courses are designed with a similar purpose. However, graduates severely lack mechanical design skills. In this study, a two-staged toolkit of a holistic flow is introduced to prevent the deficiency mentioned above. In the toolkit, mechanisms are taught by combining deductive and inductive approaches, instead of a directly inductive conventional approach. The toolkit is applied to 36 sophomore year students. Assessments of the students and their self-evaluations are collected and analysed. Findings show that the toolkit can be beneficial for teaching of mechanisms to ID students through some revisions. [ABSTRACT FROM AUTHOR]

The notion of a most specific consequence with respect to some terminological box is introduced, conditions for its existence in the description logic E L and its variants are provided, and means for its computation are developed. Algebraic properties of most specific consequences are explored. Furthermore, several applications that make use of this new notion are proposed and, in particular, it is shown how given terminological knowledge can be incorporated in existing approaches for the axiomatization of observations. For instance, a procedure for an incremental learning of concept inclusions from sequences of interpretations is developed. [ABSTRACT FROM AUTHOR]

Graph Convolutional Network (GCN) is an efficient way to deal with graph data. However, GCN aggregates node information from the downstream layers in the training and test process. In this way, GCN may expand nodes outside the current training batch, even outside the training set. In the optimization process, updating the learnable matrix involves all connected nodes in the graph convolutional operation. Therefore, learning in batched and inductive ways is unsuitable for GCN, and large-scale graph data also challenge GCN. To this end, this work proposes a Fast Multi-layer Graph Linear Network (FMLGLN) with a straightforward structure and a low amount of hyper-parameters to deal with large-scale graph data. In the implementation, FMLGLN raises the normalized adjacency matrix to the power of different values and multiplies these matrices with original features to get embedding features of nodes in pre-processing. Finally, FMLGLN concatenates these embedding features and uses a linear neural network to conduct training. The matrix multiplication is linear, and the network structure consists of linear layers, bringing a fast training speed. The features of FMLGLN are (1) batched training, (2) inductive learning, and (3) suiting large-scale graph data. Experimental results on large-scale graph data demonstrate the effectiveness of the proposed FMLGLN. [ABSTRACT FROM AUTHOR]

The need for timely information in the e-business world provides the impetus to develop a flexible database system with the capability to adapt and maintain performance levels under changing queries and changing business environments. Recognizing the importance of providing fast access to a variety of read-only applications in today's e-business world, we introduce the systems architecture for developing and implementing a flexible database system to achieve considerable gains in processing times of read queries. The key component of a flexible database system is query mining, the concept of determining relationships among query properties, alternative database structures, and query processing times. We validate the flexible database system concept through extensive laboratory experiments, where we embed learning tools to demonstrate the implementation of query mining. [ABSTRACT FROM AUTHOR]

The article studies and develops the methods for assessing the degree of participation of power plants in the general primary frequency control in a unified energy system (UES) of Russia based on time series analysis of frequency and power. To identify the processes under study, methods of associative search are proposed. The methods are based on process knowledgebase development, data mining, associative research, and inductive learning. Real-time identification models generated using these algorithms can be used in automatic control and decision support systems. Evaluation of the behavior of individual UES members enables timely prevention of abnormal and emergency situations. Methods for predictive diagnostics of generating equipment in terms of their readiness to participate in the primary frequency control are also proposed. In view of the non-stationarity of the load in electrical networks, the algorithms have been developed using wavelet analysis. Case studies are given showing the operating of the proposed methods.

Abstract Backgrounds A large number of long intergenic non-coding RNAs (lincRNAs) are linked to a broad spectrum of human diseases. The disease association with many other lincRNAs still remain as puzzle. Validation of such links between the two entities through biological experiments are expensive. However, a plethora lincRNA-data are available now, thanks to the High Throughput Sequencing (HTS) platforms, Genome Wide Association Studies (GWAS), etc, which opens the opportunity for cutting-edge machine learning and data mining approaches to extract meaningful relationships among lincRNAs and diseases. However, there are only a few in silico lincRNA-disease association inference tools available to date, and none of them utilizes side information of both the entities simultaneously in a single framework. Methods The recently developed Inductive Matrix Completion (IMC) technique provides a recommendation platform among two entities considering respective side information about them. However, the formulation of IMC is incapable of handling noise and outliers that may be present in the datasets, while data sparsity consideration is another issue with the standard IMC method. Thus, a robust version of IMC is needed that can solve the two issues. As a remedy, in this paper, we propose Stable Robust Inductive Matrix Completion (SRIMC) that utilizes the l 2,1 norm based regularization to optimize the objective function with a unique 2-step stable solution approach. Results We applied SRIMC to the available association data between human lincRNAs and OMIM disease phenotypes as well as a diverse set of side information about the lincRNAs and the diseases. The method performs better than the state-of-the-art methods in terms of p r e c i s i o n @ k and r e c a l l @ k at the top-k disease prioritization to the subject lincRNAs. We also demonstrate that SRIMC is equally effective for querying about novel lincRNAs, as well as predicting rank of a newly known disease for a set of well-characterized lincRNAs. Conclusions With the experimental results and computational evaluation, we show that SRIMC is robust in handling datasets with noise and outliers as well as dealing with novel lincRNAs and disease phenotypes.

The present study presents the instructional procedures of the implementation of corpus-aided discovery learning and the extent to which corpus-aided discovery learning can create motivating and successful English grammar learning environment for EFL (English as a Foreign Language) students. Corpora promotes inductive learning in terms of that students not only acquire grammar by language data, observation and self-discovery of rules, but also find it entertaining and exciting to make grammar rule generalization on their own (Cheng, 2011; Hunston,2002). Drawing on data of a questionnaire and an interview, the students recounted that incorporating corpus-informed data into grammar learning was very satisfying and useful, and students can be sure that the language they are practicing is up-to-date, used in everyday situations. It is not artificial or invented language but consists of the most widely used grammar (McCarthy, 2004). In addition, language corpora helped students to learn about certain language uses that were not available in any of the traditional tools. Another finding shows that students understood how certain language forms were used by native speakers of English, and students revealed more positive attitude toward corpus in learning English grammar. The findings imply that corpus-informed data can be a means of encouraging autonomous learning, illustrating real uses of abstract rules, presenting authentic daily grammar, and promoting exploratory inductive learning. However, some students, especially inexperienced students, found learning grammar through corpus-informed data was difficult or boring

State of the art algorithms for many pattern recognition problems rely on data-driven deep network models. Training these models requires a large labeled dataset and considerable computational resources. Also, it is difficult to understand the working of these learned models, limiting their use in some critical applications. Toward addressing these limitations, our architecture draws inspiration from research in cognitive systems, and integrates the principles of commonsense logical reasoning, inductive learning, and deep learning. As a motivating example of a task that requires explainable reasoning and learning, we consider Visual Question Answering in which, given an image of a scene, the objective is to answer explanatory questions about objects in the scene, their relationships, or the outcome of executing actions on these objects. In this context, our architecture uses deep networks for extracting features from images and for generating answers to queries. Between these deep networks, it embeds components for non-monotonic logical reasoning with incomplete commonsense domain knowledge, and for decision tree induction. It also incrementally learns and reasons with previously unknown constraints governing the domain's states. We evaluated the architecture in the context of datasets of simulated and real-world images, and a simulated robot computing, executing, and providing explanatory descriptions of plans and experiences during plan execution. Experimental results indicate that in comparison with an “end to end” architecture of deep networks, our architecture provides better accuracy on classification problems when the training dataset is small, comparable accuracy with larger datasets, and more accurate answers to explanatory questions. Furthermore, incremental acquisition of previously unknown constraints improves the ability to answer explanatory questions, and extending non-monotonic logical reasoning to support planning and diagnostics improves the reliability and efficiency of computing and executing plans on a simulated robot.

This paper illustrates a teaching methodology which implements some motivational mechanics of games to help overcome the widespread lack of interest of students and make learning engaging. A learning cycle that promotes three main transformative dimensions within the overall learning process is detailed: from deductive to inductive teaching; from transmissive to constructivist teaching; from summative to formative assessment. Some concrete examples of learning activities are provided. Finally, the implications of the methodology resulting from a quasi-experimental study conducted in a high school are discussed. The study compared two classes, experimental and control, in relation to the following variables: self-determination towards studying, basic psychological needs, and support for autonomy. The results show that there are significant differences in the analyzed variables, suggesting that the proposed methodology could be effective in determining positive changes in motivational dynamics.

Nonlocal phonological interactions are often sensitive to morphological domains. Bolivian Aymara restricts the cooccurrence of plain, ejective, and aspirated stops within, but not across, morphemes. We document these restrictions in a morphologically parsed corpus of Aymara. We further present two experiments with native Aymara speakers. In the first experiment, speakers are asked to repeat nonce words that should be interpreted as monomorphemic. Speakers are more accurate at repeating nonce words that respect the nonlocal phonotactic restrictions than nonce words that violate them. In a second experiment, some nonce words are interpetable as morphologically complex, while others suggest a monomorphemic parse. Speakers show a sensitivity to this difference, and repeat the words more accurately when they can be interpreted as having a morpheme boundary between two consonants that tend to not cooccur inside a morpheme. Finally, we develop a computational model that induces nonlocal representations from the baseline grammar. The model posits projections when it notices that certain segments often cooccur when separated by a morpheme boundary. The model generates a full Maximum Entropy phonotactic grammar, which makes distinctions between attested and rare/unattested sequences in a way that aligns with the speaker behavior.

Analysis of human judgment and decision making provides useful methodologies for examining the human decision process and substantive results. One such methodology is a lens model analysis. The authors used such a model to study how well a model of expert decisions can capture a valid strategy in the decision process. The study also addresses whether a model of an expert can be more accurate than the expert. The predictive accuracy (predictive validity) of two linear (statistical) models and two nonlinear models of human experts is compared. The results indicate that nonlinear models can capture factors (valid nonlinear strategy) that contribute to the experts' predictive accuracy. However, linear models cannot capture the valid nonlinear strategy as well as nonlinear models. One linear model and two nonlinear models performed as well as the overall average of a group of experts. However, all of the models were outperformed by the most accurate expert. By combining validity of decision strategy with characteristics of modeling algorithms, it is possible to explain why certain algorithms perform better than others. [ABSTRACT FROM AUTHOR]

Copyright of ESTOA: Revista de la Facultad de Arquitectura y Urbanismo de la Universidad de Cuenca is the property of ESTOA Revista de la Facultad de Arquitectura y Urbanismo de la Universidad de Cuenca and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Knowledge acquisition is the process of accumulating new information and relating it to what is already known. Knowledge acquisition has been regarded as the bottleneck in knowledge-based systems development. In this paper, a distributed knowledge acquisition system (DKAS) is introduced for automating decision rules construction from a set of examples in a decision support system. DKAS has the potential to include various learning mechanisms and employs a multi-agent and parallel processing paradigm. The implementation of a DKAS integrates inductive and deductive learning methods that use different learning strategies. A stock selection problem is used to demonstrate the effectiveness of DKAS in solving classification type problems. The performance of the DKAS in portfolio management is compared to the performance of the NYSE and the S&P 500. The results indicate that the rules derived from using the DKAS outperform both the NYSE and the S&P 500. [ABSTRACT FROM AUTHOR]

This paper presents a layered approach to inductive learning that improves the stability and accuracy of the knowledge acquisition method. The methodology is applied to international financial analysis where the comparison of risk structures of different economic environments is critical. The results demonstrate the efficacy of the learning approach for financial decision support. The highly explicit decision structures are especially useful for comparing national differences. The method's utility for decision integration is also discussed. [ABSTRACT FROM AUTHOR]

The growing amount of information in the world has increased the need for computerized classification of different objects. This situation is present in higher education as well where the possibility of effortless detection of similarity between different study courses would give the opportunity to organize student exchange programmes effectively and facilitate curriculum management and development. This area which currently relies on manual time-consuming expert activities could benefit from application of smartly adapted machine learning technologies. Data in this problem domain is complex leading to inability for automatic classification approaches to always reach the desired result in terms of classification accuracy. Therefore, our approach suggests an automated/semi-automated classification solution, which incorporates both machine learning facilities and interactive involvement of a domain expert for improving classification results. The system's prototype has been implemented and experiments are carried out. This interactive classification system allows to classify educational data, which often comes in unstructured or semi-structured, incomplete and/or insufficient form, thus reducing the number of misclassified instances significantly in comparison with the automatic machine learning approach.

The wide 3D applications have led to increasing amount of 3D object data, and thus effective 3D object classification technique has become an urgent requirement. One important and challenging task for 3D object classification is how to formulate the 3D data correlation and exploit it. Most of the previous works focus on learning optimal pairwise distance metric for object comparison, which may lose the global correlation among 3D objects. Recently, a transductive hypergraph learning has been investigated for classification, which can jointly explore the correlation among multiple objects, including both the labeled and unlabeled data. Although these methods have shown better performance, they are still limited due to 1) a considerable amount of testing data may not be available in practice and 2) the high computational cost to test new coming data. To handle this problem, considering the multi-modal representations of 3D objects in practice, we propose an inductive multi-hypergraph learning algorithm, which targets on learning an optimal projection for the multi-modal training data. In this method, all the training data are formulated in multi-hypergraph based on the features, and the inductive learning is conducted to learn the projection matrices and the optimal multi-hypergraph combination weights simultaneously. Different from the transductive learning on hypergraph, the high cost training process is off-line, and the testing process is very efficient for the inductive learning on hypergraph. We have conducted experiments on two 3D benchmarks, i.e., the NTU and the ModelNet40 data sets, and compared the proposed algorithm with the state-of-the-art methods and traditional transductive multi-hypergraph learning methods. Experimental results have demonstrated that the proposed method can achieve effective and efficient classification performance. We also note that the proposed method is a general framework and has the potential to be applied in other applications in practice. [ABSTRACT FROM AUTHOR]

The current study aimed to replicate the results of previous studies examining the spacing and testing effect by showing a benefit of spaced practice and repeated testing on inductive learning. Seventy-four participants practiced diagnosing 36 case studies of six psychological disorders and tested in a posttest phase. Although learning occurred, there were no significant differences found in posttest scores between the stimuli that were practiced in a massed versus spaced format. There were also no differences found in posttest scores between stimuli that were practiced as study versus testing trials. The results of the current study necessitate a discussion about how spacing and testing can be most effective and if the effectiveness is conditional on the material being studied. [ABSTRACT FROM AUTHOR]

There have always been debates on how to teach, what to teach and when to teach what. Over time, teaching methods have frequently been revisited and updated. Recently, with the advent of computer technology, there emerged some new teaching methodologies like corpus-based learning (also called Data-Driven Learning (DDL)) and computer-assisted language learning (CALL). This means that the practices and the attitudes towards English teaching need to be updated, too. This is not, however, the case in English classes in MENA countries where teachers predominantly use the traditional methods of teaching. Consequently, recent reports have shown that MENA countries have the very lowest proficiency in English language in the world. This paper introduces corpus linguistics and corpus-based learning to the MENA English classrooms in order to improve the English language learning process and make, for the time being, a balance between traditional and modern teaching methods. This balance can be realized in incorporating inductive learning with ductive learning one the one hand, and student-centered classes and teacher-centered classes on the other. In order to showcase the advantages of using corpus-based method in the language classroom, corpus-based materials and concordances are designed using the Sketch Engine and the Brown corpus. [ABSTRACT FROM AUTHOR]