Autonomic discovery of subgoals in hierarchical reinforcement learning

Option is a promising method to discover the hierarchical structure in reinforcement learning （RL） for learning acceleration. The key to option discovery is about how an agent can find useful subgoals autonomically among the passing trails. By analyzing the agent＇s actions in the trails, useful heuristics can be found. Not only does the agent pass subgoals more frequently, but also its effective actions are restricted in subgoals. As a consequence, the subgoals can be deemed as the most matching action-restricted states in the paths. In the grid-world environment, the concept of the unique-direction value reflecting the action-restricted property was introduced to find the most matching action-restricted states. The unique-direction-value （UDV） approach is chosen to form options offline and online autonomically. Experiments show that the approach can find subgoals correctly. Thus the Q-learning with options found on both offline and online process can accelerate learning significantly.     

School performance,social networking effects,and learning of school children: Evidence of reciprocal relationships in Abu Dhabi

This study uses structural equations modeling to test a hypothetical social network model with applications to a sample of 34,896 school children in Abu Dhabi. The main independent constructs in the model are related to children’s attitude with regard to social networking, reasons for using social networks, things done on social networks, and topics used. The dependent constructs cover perceived school performance and social effects of social networking. The study will describe the relations among the various constructs. The effect of other variables, such as parental knowhow, is also investigated. Our work has improved our insight in the social networking model. Results support the idea of reciprocal relations among perceived performance, learning from social networking, and the effect of social networking. Evidence for a model that includes opposite pathways implies that the problem of social networking constructs, its antecedents, and possible consequences should be examined with caution.     

基于集成分类的暴恐图像自动标注方法

为减少暴恐图像对社会发展和青少年成长造成的不利影响，本文提出一种基于集成分类的暴恐图像自动标注方法，辅助筛除网页中的暴恐信息。该方法将暴恐图像的标注视作多标签分类问题，利用迁移学习训练多个子网络，然后通过集成学习对子网络的输出进行融合，同时在融合过程中针对各个标签在不同网络上的准确率进行权重分配，最后经过一系列矩阵运算得到图像的标注结果。实验结果表明，与传统机器学习算法相比，本文方法在准确率和召回率上都有较大提升，并改善了样本不均衡所造成的不同标签类别上模型标注精确度差异较大的问题。     

基于深度学习的电力服务调度工单智能处理的研究

为打造以客户为中心的现代供电服务体系,进一步提升为民服务的质量和水平,强化电力保障和优质服务,贵安供电局开展海量工单分析,从而实现服务调度业务薄弱点的发现和改进。因此,提出基于深度学习的工单识别分类技术应用,通过深度学习进行建模、工单的标签特征进行提炼、并建立训练模型进行学习、对故障单和意见单进行识别,优化投诉风险预警与管理工作,缓解服务调度工作人员服务压力。     

A global perspective on an emotional learning model proposal

Existing literature argues that emotions have a significant impact on the majority of human activities and functions. The learning process is one of the activities on which emotions have a direct influence. Thus, understanding the manner in which emotions change the students’ learning process is not only very important but it can also allow to improve the existing learning models.Currently, in the majority of situations, the teacher serves as a facilitator between the student and the learning course, and through a constant analysis of the student’s behaviour, emotions and achievements, he constantly performs adjustments to the teaching process in order to meet the students’ needs and goals. Thus far, in online learning environments there is no easy way for teachers to analyse students’ behaviour and emotions. A possible solution to this problem can be the development of mechanisms that enable computers to automatically detect students’ emotions and adapt the learning process in order to meet students’ real needs.An emotional learning model was described and a software prototype was developed and tested, in order to find out whether it performs live identification of the students’ emotions, by using affective computing techniques, and whether it automatically performs adjustments to their individual learning process. Through a deeper analysis and multidisciplinary discussion of the achieved results it is possible to acknowledge that not only emotions impact students’ learning, but also that an application that performs live emotion recognition and which integrates this feature with adjustable online learning environments will trigger improvements in students’ learning.     

网络环境下智能化学习模式构建与实践研究

随着计算机技术和网络技术的飞速发展,以及各种移动设备的应用与普及,网络环境下的学习方式已经成为新时期现代化教学方式发展的新方式。但目前网络环境下的学习方式还存在如上课形式单一、学习资源不足、学生自主学习能力弱、课堂互动较少、各种教学活动不易于控制等问题,那么就急需构建网络环境下新的学习方式。将智能化融入教学过程中不仅能够改进目前网上教学所遇到的问题,而且能提升学习者的学习效果,实现教学资源个性化、教学活动情景化、教学管理智能化。     

Avoiding collaborative paradox in multi-agent reinforcement learning

The collaboration productively interacting between multi-agents has become an emerging issue in real-world applications. In reinforcement learning, multi-agent environments present challenges beyond tractable issues in single-agent settings. This collaborative environment has the following highly complex attributes: sparse rewards for task completion, limited communications between each other, and only partial observations. In particular, adjustments in an agent's action policy result in a nonstationary environment from the other agent's perspective, which causes high variance in the learned policies and prevents the direct use of reinforcement learning approaches. Unexpected social loafing caused by high dispersion makes it difficult for all agents to succeed in collaborative tasks. Therefore, we address a paradox caused by the social loafing to significantly reduce total returns after a certain timestep of multi-agent reinforcement learning. We further demonstrate that the collaborative paradox in multi-agent environments can be avoided by our proposed effective early stop method leveraging a metric for social loafing.     

基于智慧学习的高职信息技术教学活动设计

学术界一般认为所谓智慧教育其实就是一种新的教育形态,主要体现为依托新兴的互联网技术与创新型教学方法及教学内容,对知识进行全方位的联合,从而通过优化教学质量、更新教育体系、重构教学模式等途径来完成智慧转型和升级。文章探讨了智慧学习环境概念以及技术特征,分析了智慧学习环境下高职信息技术活动设计存在的问题,研究了智慧学习环境下高职信息技术教学活动设计的方法和措施。     

Multi-view representation learning for multi-view action recognition

Although multiple methods have been proposed for human action recognition, the existing multi-view approaches cannot well discover meaningful relationship among multiple action categories from different views. To handle this problem, this paper proposes an multi-view learning approach for multi-view action recognition. First, the proposed method leverages the popular visual representation method, bag-of-visual-words (BoVW)/fisher vector (FV), to represent individual videos in each view. Second, the sparse coding algorithm is utilized to transfer the low-level features of various views into the discriminative and high-level semantics space. Third, we employ the multi-task learning (MTL) approach for joint action modeling and discovery of latent relationship among different action categories. The extensive experimental results on M²I and IXMAS datasets have demonstrated the effectiveness of our proposed approach. Moreover, the experiments further demonstrate that the discovered latent relationship can benefit multi-view model learning to augment the performance of action recognition.     

Diagnosis of learner dropout based on learning styles for online distance learning

The amount of data generated by computer systems in Online Distance Learning (ODL) contains rich information. One example of this information we define as the Learner Learning Trail (LLT), which is the sequence of interactions between the students and the virtual environment. Another example is the Learner Learning Style (LLS), which is associated with the student behavior and choices during the learning process. This information can be used to identify learner behavior and learning style. We perceived, after the study of related literature, that the research field of learner diagnosis for ODL does not apply the conjoint use of LLT and LLS. In this article, we propose a model capable of integrating data generated from the behavior of students in ODL with cognitive aspects of them, such as their Learning Styles, by crossing LLT with LLS. We also propose the CPAD method (Collect, Preprocessing, Analysis, Diagnosis), which is implemented by collecting the raw data regarding learning activities, preprocessing the data into structured time sequences, analyzing the sequences regarding the learning styles and using this analysis to diagnose the learner behavior. We selected the dropout to investigate, once the dropout rate in ODL is a real problem in universities around the world. In addition, the dropout is a student decision which can be associated with previous students behaviors. We performed a study with 202 learners to evaluate if learning styles are capable of explaining aspects of the student behavior. The results suggest that Sequential/Global learning style dimension is more capable of explaining the dropout than the other dimensions. Also, we performed four classification experiments to verify how the dimensions of Felder-Silverman Learning Style Model influence the learner diagnosis. We perceived that the Sequential/Global dimension could provide a higher accuracy average with lower variation independently of the diagnosis technique.     

Experiential learning prepares students to assume professionalroles

The technical communication program in the English Department at Oklahoma State University is described. The author describes how teachers can structure course work so the students complete assignments similar to projects they will complete as professionals. To provide experiential learning outside the classroom, it is suggested that educators include internships or cooperative work-study opportunities in their academic programs, encourage students to become active members in professional organizations, and help students complete professional activities such as giving papers at meetings and conferences and publishing articles in newsletters, conference proceedings, and journals     

Designing and managing virtual learning communities

There is a tendency to view education on the Internet as simply a more efficient way to access information and to communicate, but the Internet is much more than just another tool. The Internet has the potential to create communities where students participate in robust discourse and rituals of communication, establish their identities, and traverse community boundaries. We believe we need to design on-line courses with sound pedagogical frameworks and with a sense of promoting community values of diversity, connectedness, and civic responsibility. Therefore, the purpose of this paper is to describe a framework that we use to design virtual learning communities. We explain community activities to consider, describe how we used our framework for designing three classes, and pose issues that arose when using this framework. We hope our thoughts will direct discussion toward the creation of innovative learning communities     

An incremental learning method based on probabilistic neural networks and adjustable fuzzy clustering for human activity recognition by using wearable sensors

Human activity recognition by using wearable sensors has gained tremendous interest in recent years among a range of health-related areas. To automatically recognize various human activities from wearable sensor data, many classification methods have been tried in prior studies, but most of them lack the incremental learning abilities. In this study, an incremental learning method is proposed for sensor-based human activity recognition. The proposed method is designed based on probabilistic neural networks and an adjustable fuzzy clustering algorithm. The proposed method may achieve the following features. 1) It can easily learn additional information from new training data to improve the recognition accuracy. 2) It can freely add new activities to be detected, as well as remove existing activities. 3) The updating process from new training data does not require previously used training data. An experiment was performed to collect realistic wearable sensor data from a range of activities of daily life. The experimental results showed that the proposed method achieved a good tradeoff between incremental learning ability and the recognition accuracy. The experimental results from comparison with other classification methods demonstrated the effectiveness of the proposed method further.     

Efficient web-based open and distance learning services

In this paper, we present data management issues faced during the design and development of an open distance learning system for the University of Patras, Greece. In order to handle data efficiently, as required in a web tele-training application, for each type of information maintained, different strategies must be deployed according to their behaviour and structure. The diversity and complexity of data, the network aspect of the application and web deficiencies impose an architecture design incorporating a plethora of technologies and tools that must be integrated in such a fashion that they efficiently organise these data preserving their relationships. This presents a software engineering challenge requiring coherence of solutions at all levels: structures, consistency, security, models, and protocols. The paper presents the data components of an open and distance learning (ODL) system that access the information stored in a database and the file system, their underlying technology, their interaction with the network services, and features regarding the ways they address issues faced in an open vendor-independent distance learning environment and outlines the system's overall architecture. In addition, this paper presents the architecture, the design and the services of a network-based information system that supports open and distance learning activities. The open and distance learning information system (ODLIS) offers synchronous and asynchronous distance learning and management of information system (MIS) services to support the educational procedure. The ODLIS is a web-based application, which runs over the Internet using real time protocols.     

Keyed learning: An adversarial learning framework—formalization,challenges, and anomaly detection applications

We propose a general framework for keyed learning, where a secret key is used as an additional input of an adversarial learning system. We also define models and formal challenges for an adversary who knows the learning algorithm and its input data but has no access to the key value. This adversarial learning framework is subsequently applied to a more specific context of anomaly detection, where the secret key finds additional practical uses and guides the entire learning and alarm‐generating procedure.     

Green learning: Introduction,examples and outlook

Rapid advances in artificial intelligence (AI) in the last decade have been largely built upon the wide applications of deep learning (DL). However, the high carbon footprint yielded by larger and larger DL networks has become a concern for sustainability. Furthermore, DL decision mechanism is somewhat obscure in that it can only be verified by test data. Green learning (GL) is being proposed as an alternative paradigm to address these concerns. GL is characterized by low carbon footprints, lightweight model, low computational complexity, and logical transparency. It offers energy-efficient solutions in cloud centers as well as mobile/edge devices. GL also provides a more transparent, logical decision-making process which is essential to gaining people’s trust. Several statistical tools such as unsupervised representation learning, supervised feature learning, and supervised decision learning, have been developed to achieve this goal in recent years. We have seen a few successful GL examples with performance comparable with state-of-the-art DL solutions. This paper introduces the key characteristics of GL, its demonstrated applications, and future outlook.     

Securing Internet of Things (IoT) with machine learning

Advances in hardware, software, communication, embedding computing technologies along with their decreasing costs and increasing performance have led to the emergence of the Internet of Things (IoT) paradigm. Today, several billions of Internet‐connected devices are part of the IoT ecosystem. IoT devices have become an integral part of the information and communication technology (ICT) infrastructure that supports many of our daily activities. The security of these IoT devices has been receiving a lot of attention in recent years. Another major recent trend is the amount of data that is being produced every day which has reignited interest in technologies such as machine learning and artificial intelligence. We investigate the potential of machine learning techniques in enhancing the security of IoT devices. We focus on the deployment of supervised, unsupervised learning techniques, and reinforcement learning for both host‐based and network‐based security solutions in the IoT environment. Finally, we discuss some of the challenges of machine learning techniques that need to be addressed in order to effectively implement and deploy them so that they can better protect IoT devices.     

A reinforcement learning ticket-based probing path discovery scheme for MANETs

In this paper, a path discovery scheme which supports QoS routing in mobile ad hoc networks (MANETs) in the presence of imprecise information is investigated. The aim is to increase the probability of success in finding feasible paths and reduce average path cost of a previously proposed ticket based probing (TBP) path discovery scheme. The proposed scheme integrates the original TBP scheme with a reinforcement learning method called the on-policy first-visit Monte Carlo (ONMC) method. We investigate the performance of the ONMC method in the presence of imprecise information. Our numerical study shows that, in respect to a flooding based algorithm, message overhead reduction can be achieved with marginal difference in the path search ability and additional computational and storage requirements. When the average message overhead of the ONMC method is reduced to the same order of magnitude of the original TBP, the ONMC method gains an improvement of 28% in success ratio and 7% reduction in the average path cost over the original TBP.     

使用深度学习的社交物联网信任预测模型

社交物联网是社交网络概念在物联网中整合后兴起的一个蓬勃发展的研究领域.提出了一种适用于社交物联网网络的改进型节点级信任模型,并通过与其他信任模型的对比仿真实验证明在恶意节点的攻击下,提出的模型拥有更好的稳定性和适用性,总体波动较小.同时,针对实际社交物联网网络中新加入网络的陌生节点可能遇到的网络延迟影响信任值评估的问题...     

E—Learning系统的个性化推荐方法研究

以E—Learning系统建设为背景,通过采用文献查找、调查研究等方法探讨个性化推荐理论的内涵,并结合当前建设中的E—Learning系统,分析了目前常用的个性化推荐策略,并进行介绍比较和分析以后,总结经验,以应用于E—Learning系统的建设。提出适合于E—learning系统建设的个性化推荐策略：采用关联规则推荐策略和协同过滤技术,基于WEB技术建立一个虚拟学习系统,利用推荐算法,结合用户需求,将学习的资源、学习活动和学习策略进行整合,向用户推荐完整的满足用户需求的E—Learning学习方案。