scholarly journals MSGCN: Multi-Subgraph Based Heterogeneous Graph Convolution Network Embedding

2021 ◽  
Vol 11 (21) ◽  
pp. 9832
Author(s):  
Junhui Chen ◽  
Feihu Huang ◽  
Jian Peng
Keyword(s):  
Sampling Method ◽  
Inductive Learning ◽  
Auxiliary Information ◽  
Graph Embedding ◽  
Learning Task ◽  
Network Embedding ◽  
Transductive Learning ◽  
Learning Tasks ◽  
Topology Information ◽  
Important Nodes

Heterogeneous graph embedding has become a hot topic in network embedding in recent years and has been widely used in lots of practical scenarios. However, most of the existing heterogeneous graph embedding methods cannot make full use of all the auxiliary information. So we proposed a new method called Multi-Subgraph based Graph Convolution Network (MSGCN), which uses topology information, semantic information, and node feature information to learn node embedding vector. In MSGCN, the graph is firstly decomposed into multiple subgraphs according to the type of edges. Then convolution operation is adopted for each subgraph to obtain the node representations of each subgraph. Finally, the node representations are obtained by aggregating the representation vectors of nodes in each subgraph. Furthermore, we discussed the application of MSGCN with respect to a transductive learning task and inductive learning task, respectively. A node sampling method for inductive learning tasks to obtain representations of new nodes is proposed. This sampling method uses the attention mechanism to find important nodes and then assigns different weights to different nodes during aggregation. We conducted an experiment on three datasets. The experimental results indicate that our MSGCN outperforms the state-of-the-art methods in multi-class node classification tasks.

A Survey of Network Embedding for Drug Analysis and Prediction

2020 ◽  
Vol 21 ◽  
Author(s):  
Zhixian Liu ◽  
Qingfeng Chen ◽  
Wei Lan ◽  
Jiahai Liang ◽  
Yiping Pheobe Chen ◽  
...  
Keyword(s):  
Deep Learning ◽  
Protein Function ◽  
Dimensional Space ◽  
Auxiliary Information ◽  
Matrix Decomposition ◽  
Drug Analysis ◽  
Machine Learning Algorithms ◽  
Superior Performance ◽  
Network Embedding ◽  
Similarity Estimation

: Traditional network-based computational methods have shown good results in drug analysis and prediction. However, these methods are time consuming and lack universality, and it is difficult to exploit the auxiliary information of nodes and edges. Network embedding provides a promising way for alleviating the above problems by transforming network into a low-dimensional space while preserving network structure and auxiliary information. This thus facilitates the application of machine learning algorithms for subsequent processing. Network embedding has been introduced into drug analysis and prediction in the last few years, and has shown superior performance over traditional methods. However, there is no systematic review of this issue. This article offers a comprehensive survey of the primary network embedding methods and their applications in drug analysis and prediction. The network embedding technologies applied in homogeneous network and heterogeneous network are investigated and compared, including matrix decomposition, random walk, and deep learning. Especially, the Graph neural network (GNN) methods in deep learning are highlighted. Further, the applications of network embedding in drug similarity estimation, drug-target interaction prediction, adverse drug reactions prediction, protein function and therapeutic peptides prediction are discussed. Several future potential research directions are also discussed.

scholarly journals TPOT-NN: augmenting tree-based automated machine learning with neural network estimators

2021 ◽  
Author(s):  
Joseph D. Romano ◽  
Trang T. Le ◽  
Weixuan Fu ◽  
Jason H. Moore
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Binary Classification ◽  
Inductive Learning ◽  
Future Directions ◽  
High Performing ◽  
Learning Tasks ◽  
Benchmark Datasets ◽  
Automated Machine Learning ◽  
Standard Tree

AbstractAutomated machine learning (AutoML) and artificial neural networks (ANNs) have revolutionized the field of artificial intelligence by yielding incredibly high-performing models to solve a myriad of inductive learning tasks. In spite of their successes, little guidance exists on when to use one versus the other. Furthermore, relatively few tools exist that allow the integration of both AutoML and ANNs in the same analysis to yield results combining both of their strengths. Here, we present TPOT-NN—a new extension to the tree-based AutoML software TPOT—and use it to explore the behavior of automated machine learning augmented with neural network estimators (AutoML+NN), particularly when compared to non-NN AutoML in the context of simple binary classification on a number of public benchmark datasets. Our observations suggest that TPOT-NN is an effective tool that achieves greater classification accuracy than standard tree-based AutoML on some datasets, with no loss in accuracy on others. We also provide preliminary guidelines for performing AutoML+NN analyses, and recommend possible future directions for AutoML+NN methods research, especially in the context of TPOT.

scholarly journals Achieving 90% In Data-Centric Industry Deep Learning Task

2021 ◽  
Author(s):  
Tong Guo
Keyword(s):  
Deep Learning ◽  
Object Detection ◽  
Noisy Data ◽  
Learning Task ◽  
Simple Method ◽  
Sequence Generation ◽  
Learning Tasks ◽  
Human Evaluation ◽  
Model Predictions ◽  
Click Through Rate

In industry deep learning application, our manually labeled data has a certain number of noisy data. To solve this problem and achieve more than 90 score in dev dataset, we present a simple method to find the noisy data and re-label the noisy data by human, given the model predictions as references in human labeling. In this paper, we illustrate our idea for a broad set of deep learning tasks, includes classification, sequence tagging, object detection, sequence generation, click-through rate prediction. The experimental results and human evaluation results verify our idea.

scholarly journals Designing professional learning tasks for mathematics learning trajectories

2013 ◽  
Vol 7 (4) ◽  
pp. 135-143
Author(s):  
P. Holt Wilson ◽  
Paola Sztajn ◽  
Cyndi Edgington
Keyword(s):  
Elementary Teachers ◽  
Teacher Learning ◽  
Teacher Knowledge ◽  
Professional Learning ◽  
Mathematics Learning ◽  
Theoretical Perspective ◽  
Learning Task ◽  
Learning Trajectory ◽  
Learning Trajectories ◽  
Learning Tasks

In this paper, we present an emerging set of learning conjectures and design principles to be used in the development of professional learning tasks that support elementary teachers’ learning of mathematics learning trajectories. We outline our theoretical perspective on teacher knowledge of learning trajectories, review the literature concerning mathematics professional learning tasks, offer a set of initial conjectures about teacher learning of learning trajectories, and articulate a set of principles to guide the design of tasks. We conclude with an example of one learning trajectory professional learning task taken from our current research project.Diseño de tareas de aprendizaje profesional para trayectorias de aprendizaje de matemáticasEn este artículo, presentamos un conjunto emergente de conjeturas de aprendizaje y de principios de diseño para ser empleados en el desarrollo de tareas de aprendizaje profesional que apoyan el aprendizaje de trayectorias de aprendizaje de matemáticas de maestros de primaria. Describimos brevemente nuestra perspectiva teórica sobre el conocimiento del profesor acerca de trayectorias de aprendizaje; revisamos la literatura sobre tareas de aprendizaje profesional, presentamos un conjunto de conjeturas iniciales acerca del aprendizaje del profesor sobre trayectorias de aprendizaje; y articulamos un conjunto de principios para guiar el diseño de tareas. Concluimos con un ejemplo de una tarea de aprendizaje profesional que ha sido tomada de nuestro proyecto de investigación actual.Handle: http://hdl.handle.net/10481/24791Nº de citas en WOS (2017): 3 (Citas de 2º orden, 2)Nº de citas en SCOPUS (2017): 3 (Citas de 2º orden, 1)

scholarly journals JANE: Jointly Adversarial Network Embedding

2020 ◽  
Author(s):  
Liang Yang ◽  
Yuexue Wang ◽  
Junhua Gu ◽  
Chuan Wang ◽  
Xiaochun Cao ◽  
...  
Keyword(s):  
Link Prediction ◽  
Real Data ◽  
Semantic Space ◽  
Network Embedding ◽  
Generative Adversarial Network ◽  
Adversarial Learning ◽  
Adversarial Network ◽  
Node Clustering ◽  
Topology Information ◽  
Embedding Methods

Motivated by the capability of Generative Adversarial Network on exploring the latent semantic space and capturing semantic variations in the data distribution, adversarial learning has been adopted in network embedding to improve the robustness. However, this important ability is lost in existing adversarially regularized network embedding methods, because their embedding results are directly compared to the samples drawn from perturbation (Gaussian) distribution without any rectification from real data. To overcome this vital issue, a novel Joint Adversarial Network Embedding (JANE) framework is proposed to jointly distinguish the real and fake combinations of the embeddings, topology information and node features. JANE contains three pluggable components, Embedding module, Generator module and Discriminator module. The overall objective function of JANE is defined in a min-max form, which can be optimized via alternating stochastic gradient. Extensive experiments demonstrate the remarkable superiority of the proposed JANE on link prediction (3% gains in both AUC and AP) and node clustering (5% gain in F1 score).

4. Adaptive Gaze Control in an Observational Learning Task

2016 ◽  
Author(s):  
James Kim
Keyword(s):  
Eye Movement ◽  
Observational Learning ◽  
Eye Gaze ◽  
Relevant Information ◽  
Learning Task ◽  
Gaze Control ◽  
Visually Guided ◽  
Object Condition ◽  
Learning Tasks ◽  
Different Types

The purpose of this study was to examine factors that influence how people look at objects they will have to act upon while watching others interact with them first. We investigated whether including different types of task-relevant information into an observational learning task would result in participants adapting their gaze towards an object with more task-relevant information. The participant watched an actor simultaneously lift and replace two objects with two hands then was cued to lift one of the two objects. The objects had the potential to change weight between each trial. In our cue condition, participants were cued to lift one of the objects every single time. In our object condition, the participants were cued equally to act on both objects; however, the weights of only one of the objects would have the potential to change. The hypothesis in the cue condition was that the participant would look significantly more at the object being cued. The hypothesis for the object condition was that the participant would look significantly more (i.e. adapt their gaze) at the object changing weight. The rationale behind this is that participants will learn to allocate their gaze significantly more towards that object so they can gain information about its properties (i.e. weight change). Pending results will indicate whether or not this occurred, and has implications for understanding eye movement sequences in visually guided behaviour tasks. The outcome of this study also has implications for the mechanisms of eye gaze with respect to social learning tasks. 

Learning Objects, Learning Tasks, and Handhelds

2011 ◽  
pp. 451-469 ◽  
Author(s):  
Daniel Churchill ◽  
John Gordon Hedberg
Keyword(s):  
Student Learning ◽  
Emerging Technologies ◽  
Main Idea ◽  
Learning Task ◽  
Learning Objects ◽  
Learning Activities ◽  
Learning Object ◽  
Digital Resources ◽  
Learning Tasks

The main idea behind learning objects is that they are to exist as digital resources separated from the learning task in which they are used. This allows a learning object to be reused with different learning tasks. However, not all learning objects operate in similar ways, neither are all learning tasks the same, and this exposes the problem that current recommendations from literature fail to link learning objects and their reuse in varied learning tasks. In this chapter, we explore definitions of learning objects and learning tasks. We also suggest that appropriate matches would lead to more effective pedagogical applications that can be used as set of recommendations for designers of learning objects and teachers who plan learning tasks and select learning objects for student learning activities. In addition, we discuss applications of learning objects delivered by emerging technologies which may change how digital resources are accessed and used by students in and out of classrooms.

A Supplantation Model for Instructional Design: Investigation of a Behavioural Science Approach

1978 ◽  
Vol 22 (3) ◽  
pp. 277-294 ◽  
Author(s):  
Lynna J. Ausburn ◽  
Floyd B. Ausburn
Keyword(s):  
Instructional Design ◽  
Cognitive Processes ◽  
Task Analysis ◽  
Learning Task ◽  
Performance Outcomes ◽  
Behavioural Science ◽  
Learning Tasks ◽  
Task Requirements ◽  
Treatment Interaction ◽  
Aptitude Treatment Interaction

Drawing on concepts from such areas as information processing and cognitive processes in learning, learning task analysis, and interactive research techniques, this paper discusses a model for instructional design which is intended to improve the reliability and predictability of the design process. The model stresses interactions among specific combinations of learning task requirements, learner characteristics, and instructional treatment properties, in a manner analogous to the well-known aptitude-treatment interaction (ATI) model. The process underlying the model is “supplantation”, which is the provision of overt assistance to learners in performing a specific process required by a task. The supplantation model for instructional design is presented as an instrument for helping to produce predictable performance outcomes through the analysis of learners and learning tasks, and the joining of learners and tasks through the use of instructional treatments which assist learners in performing task requirements.

scholarly journals Evaluation of the Consonant Vowel Consonant (CVC) Test as an Assessment and Screening Instrument for Mild Cognitive Impairment

2015 ◽  
Vol 7 (4) ◽  
pp. 119 ◽  
Author(s):  
Esther Vierck ◽  
Richard J. Porter ◽  
Janet K. Spittlehouse ◽  
Peter R. Joyce
Keyword(s):  
Cognitive Impairment ◽  
Mild Cognitive Impairment ◽  
Psychometric Properties ◽  
Word Learning ◽  
Verbal Memory ◽  
Verbal Learning ◽  
Learning Task ◽  
Screening Instrument ◽  
Learning Tasks ◽  
Ceiling Effects

<p>Objective: Traditional word learning tasks have been criticised for being affected by ceiling effects. The Consonant Vowel Consonant (CVC) test is a non-word verbal learning task designed to be more difficult and therefore have a lower risk of ceiling effects.</p><p>Method: The current study examines the psychometric properties of the CVC in 404 middle-aged persons and evaluates it as a screening instrument for mild cognitive impairment by comparing it to the Montreal Cognitive Assessment (MoCA). Differences between currently depressed and non-depressed participants were also examined.</p><p>Results: CVC characteristics are similar to traditional verbal memory tasks but with reduced likelihood of a ceiling effect. Using the standard cut-off on the MoCA as an indication of mild cognitive impairment, the CVC performed only moderately well in predicting this. Depressed participants scored significantly lower on the CVC compared with non-depressed individuals.</p><p>Conclusions: The CVC may be similar in psychometric properties to the traditional word learning tests but with a higher ceiling. Scores are lower in depression.</p>

A prioriassessment of language learning tasks by practitioners

2009 ◽  
Vol 42 (4) ◽  
pp. 504-518 ◽  
Author(s):  
Gerard J. Westhoff
Keyword(s):  
Teacher Education ◽  
Cognitive Psychology ◽  
Language Learning ◽  
A Priori ◽  
Reliability And Validity ◽  
Learning Task ◽  
Theoretical Framework ◽  
Learning Materials ◽  
Learning Tasks ◽  
Depth Analysis

Teachers' competence to estimate the effectiveness of learning materials is important and often neglected in programmes for teacher education. In this lecture I will try to explore the possibilities of designing scaffolding instruments fora prioriassessment of language learning tasks, based on insights from SLA and cognitive psychology, more specifically connectionist theory. I will subsequently outline the development and evaluation of a ‘yardstick’ to judge complex, integrated, life-like tasks, such as WebQuests. The possibilities will be explored of performing in-deptha prioritask analyses as a learning task for teachers in order to enhance their competence in making ‘educated guesses’ about task effectiveness. Finally, an experiment will be described to determine the reliability and validity of an instrument for in-depth analysis of language learning tasks based on the theoretical framework previously described.

Sign in / Sign up

Export Citation Format

Share Document