Gregory W. Hislop · 2002
Asynchronous learning networks (ALN) provide a means to break the traditional time and place constraints in delivery of education. This application of information technology to education creates exciting possibilities for extending educational opportunities. Drexel University is using an ALN to deliver a graduate degree in information systems with a software engineering concentration. Students in this program never attend a face-to-face class but nonetheless do their work in an environment that emphasizes collaborative learning. This paper provides an overview of the Drexel project, describes the learning environment, and discusses methods and issues for teaching courses in an ALN environment.
Cited by 2
Cairn.info · 2010
Cited by 0
H. A. Latchman, Ch. Salzmann, S. Thottapilly et al. · 1998
Abstract The rapid growth of the Internet and the media-rich extensions of the World Wide Web allow new developments in the way instructors transfer knowledge to the students. There is no doubt that nothing will replace student participation in an interactive class but it is sometimes not feasible for students to attend the class due to distance or time constraints. This paper documents the development of a test bed at the University of Florida for solutions to overcome both these constraints. It emphasizes a solution which allows the student to attend the class in real time from a remote place and also facilitates the simultaneously archiving of the class session and associated material for asynchronous access. The proposed solution permits live and asynchronous interactions among students and between the students and the instructor, which are crucial for effective learning. This paper describes the technical issues involved and the chosen solutions to provide enhanced live and archived classes. This paper also focuses on the evaluation of experiments conducted at the University of Florida and in some other universities and explores possible future trends to improve the proposed learning environment. 1.
Cited by 0
Nuldén, Urban, Hardless, Christian, Lundin, Johan · Chalmers Research (Chalmers University of Technology) · 2001
Cited by 0
JINR Document Server (Joint Institute for Nuclear Research)
Cited by 0
Gregory W. Hislop · Journal of the Association for Information Systems · 1997
Drexel University is delivering a graduate degree in information systems by asynchronous learning network (ALN). Students in this program never attend a face-to-face class. This paper discusses methods of structuring course material for delivery in this environment
Cited by 6Open access
n.n. · 2009
The negotiation of what is to count as mutually acceptable collaborative knowledge is difficult to conduct when participants cannot interact face-to-face. We review certain related work on negotiation support and develop a concept of “knowledge negotiation” that is appropriate for collaborative learning in ALNs used in school classrooms. This concept is situated within the framework of collaborative knowledge building viewed at the group unit of analysis; it contrasts with negotiation as the reconciliation of multiple personal opinions through voting. We then describe an implementation of support for knowledge negotiation in an ALN that is currently being tested in European schools.
Cited by 0
Christian Hardless Johan, Johan Lundin, Urban Nulden · 2001
In this paper, we report on our experiences with an asynchronous learning network (ALN) based course in higher education in which participation was not an option but a requirement. Mandatory participation in collaborative learning was the primary examination that had to be passed in order to qualify for the secondary examination, the authoring of a short essay. The evaluations show that mandatory participation as examination (MPE) is a viable format of examination for ALN based learning. The main strength of this format of examination is that it promotes active participation, which is a main ingredient in the desired learning process. Surprisingly in practice, ALN based learning with MPE was not considered beneficial for learning by the students. This is explained by the problematic shift from teacher-centered to learner-centered education and the traditional view of examination as separated from learning.
Cited by 0
Gerry Stahl · 2003
The negotiation of what is to count as mutually acceptable collaborative knowledge is difficult to conduct when participants cannot interact face-to-face. We review certain related work on negotiation support and develop a concept of "knowledge negotiation " that is appropriate for collaborative learning in ALNs used in school classrooms. This concept is situated within the framework of collaborative knowledge building viewed at the group unit of analysis; it contrasts with negotiation as the reconciliation of multiple personal opinions through voting. We then describe an implementation of support for knowledge negotiation in an ALN that is currently being tested in European schools.
Cited by 25
Starr Roxanne Hiltz, Ricki Goldman · 2004
Contents: Preface. Part I: Foundations of Research on Learning Networks. S.R. Hiltz, R. Goldman, What Are Asynchronous Learning Networks? R. Benbunan-Fich, S.R. Hiltz, L. Harasim, The Online Interaction Learning Model: An Integrated Theoretical Framework for Learning Networks. J. Fjermestad, S.R. Hiltz, Y. Zhang, Effectiveness for Students: Comparisons of In-Seat and ALN Courses. J.B. Arbaugh, S.R. Hiltz, Improving Quantitative Research on ALN Effectiveness. R. Goldman, M. Crosby, K. Swan, P. Shea, Qualitative and Quisitive Research Methods for Describing Online Learning. Part II: Learning Networks: What We Know and What We Need to Know. J.B. Arbaugh, R. Benbunan-Fich, Contextual Factors That Influence ALN Effectiveness. S.R. Hiltz, P. Shea, The Student in the Online Classroom. C. Dzuiban, P. Shea, J.B. Arbaugh, Faculty Roles and Satisfaction in Asynchronous Learning Networks. M. Alavi, D. Dufner, Technology-Mediated Collaborative Learning: A Research Perspective. R.E. Rice, S.R. Hiltz, D.H. Spencer, Media Mixes and Learning Networks. K. Swan, P. Shea, The Development of Virtual Learning Communities. R. Goldman, S.R. Hiltz, Asynchronous Learning Networks: Looking Back and Looking Forward.
Cited by 191
James Kostas · ScholarWorks@UMassAmherst (University of Massachusetts Amherst) · 2025
Reinforcement learning (RL) is a machine learning paradigm that studies how to optimize an algorithm's interaction with an environment over time based on a reward signal. In recent years, RL research has generated many ground-breaking results for simulated domains where data is plentiful, such as simulated robotics problems or game-playing [34] [44] [19] [46] [62]. However, these methods all rely on backpropagation-based deep learning (BDL), that is, deep neural networks that are (typically) trained with backpropagation. Despite its successes, BDL has significant limitations. For example, 1) it is brittle and prone to overfitting or divergence (particularly for RL), 2) it typically requires specialized computing hardware, 3) it is arguably biologically implausible [55] and 4) the network components must be differentiable and must be calculated in a specific order. As a result of its limitations, despite the successes in simulated environments where data is plentiful, deep-learning-based RL (DRL) has not resulted in many recent breakthroughs in real-world RL applications such as robotics, healthcare systems, recommender systems, and automated vehicles. Inspired both by the success of DRL and by its limitations, we study a connectionist alternative based on stochastic neural networks (SNNs); SNNs are networks where the outputs of some units are not deterministic functions of the units' inputs. The class of SNNs we study is called coagent networks [56]; coagent networks are SSNs designed for RL. Coagent networks are less constrained or unconstrained by the DRL limitations above. For example, coagent networks 1) may be more biologically plausible than deep networks trained with backpropagation, 2) may be computed in a fully-distributed asynchronous fashion, and 3) can be constructed from types of components that cannot be used in BDL. Additionally, since our approach generalizes the theoretical foundations that underlie a large class of DRL algorithms, coagent networks can even include BDL networks as components of the larger coagent network; this fact may allow practitioners to leverage the advantages of both approaches simultaneously. In this work, we show that coagent networks may be trained and run asynchronously (that is, the units do not execute simultaneously or at the same rate), and we give principled, theoretically-grounded learning rules for training these networks in this setting. We prove that asynchronous coagent networks generalize a popular class of RL algorithms known as option-critic algorithms[3], and show how the theory of coagent networks can simplify the creation and analysis of option-critic algorithms. Finally, we build on prior work to extend the theory and increase the practical capabilities of coagent networks. We extend and generalize the theory of asynchronous coagent networks to address the case where parameters are shared within the network. We demonstrate that coagent networks are capable of scaling to complex high-dimensional domains, including a challenging real-world online robotics setting.
Cited by 0Open access
Zhongyu Wang, Zhaoyang Zhang, Yuqing Tian et al. · IEEE Transactions on Wireless Communications · 2022
The conventional federated learning (FL) framework usually assumes synchronous reception and fusion of all the local models at the central aggregator and synchronous updating and training of the global model at all the agents as well. However, in a wireless network, due to limited radio resource, inevitable transmission failures and heterogeneous computing capacity, it is very hard to realize strict synchronization among all the involved user equipments (UEs). In this paper, we propose a novel asynchronous FL framework, which well adapts to the heterogeneity of users, communication environments and learning tasks, by considering both the possible delays in training and uploading the local models and the resultant staleness among the received models that has heavy impact on the global model fusion. A novel centralized fusion algorithm is designed to determine the fusion weight during the global update, which aims to make full use of the fresh information contained in the uploaded local models while avoiding the biased convergence by enforcing the impact of each UE’s local dataset to be proportional to its sample share. Numerical experiments validate that the proposed asynchronous FL framework can achieve fast and smooth convergence and enhance the training efficiency significantly.
Cited by 117
Xiaolan Liu, Jackson Ross, Yue Liu et al. · 2023
The future wireless networks are expected to support more artificial intelligence (AI)-enabled applications, such as Metaverse services, at the network edge. The AI algorithms, like deep learning, play an important role in extracting important information from a large dataset, but conventional centralized learning requires collecting the datasets that are distributed over the users and always include their personal information. Federated learning (FL) has been widely investigated to address those issues by performing learning in a distributed manner. However, it shows performance degradation for heterogeneous networks. In this paper, we introduce asynchronous and personalized FL to address the heterogeneity from different aspects. We first propose a semi-asynchronous FL (Semi-Async-FL) by adding time lag to distributed global model and enabling aggregation while receiving a small set of users. Specifically, we propose a new asynchronous-based personalized FL (Async-PFL) algorithm by considering the staleness of the personalized models in classic personalized FL. The simulations show that our proposed Async-PFL achieves better learning performance than Semi-Async-FL and personalized FL.
Cited by 1
Starr Roxanne Hiltz · 1997
New Jersey Institute of Technology has been delivering college courses via an Asynchronous Learning Network (ALN) system called the Virtual ClassroomTM for a decade, using various media mixes. Currently, two complete undergraduate degree programs are available via a mix of video plus Virtual Classroom, the B.A. in Information Systems and the B.S. in Computer Science. This paper presents preliminary findings about impacts on students, and touches on some issues and potential impacts for faculty, individual universities, and the structure of higher education. Overall ratings of courses by students who complete ALN based courses are equal or superior to those for traditional courses. Dropout or Incomplete outcomes are somewhat more prevalent, while grade distributions for those who complete tend to be similar to those for traditional courses. For both students and faculty, more startup time devoted to solving the "logistics " of ALN delivery seems to be required at the beginning of courses. ALN delivery is not just a "different " way of doing the same thing, however; it is likely to change the nature and structure of higher education. KEYWORDS virtual classroom aln collaborative learning computer mediated communication computers and education I.
Cited by 0
Lukas Faber, Roger Wattenhofer · arXiv (Cornell University) · 2022
This paper studies asynchronous message passing (AMP), a new paradigm for applying neural network based learning to graphs. Existing graph neural networks use the synchronous distributed computing model and aggregate their neighbors in each round, which causes problems such as oversmoothing and limits their expressiveness. On the other hand, AMP is based on the asynchronous model, where nodes react to messages of their neighbors individually. We prove that (i) AMP can simulate synchronous GNNs and that (ii) AMP can theoretically distinguish any pair of graphs. We experimentally validate AMP's expressiveness. Further, we show that AMP might be better suited to propagate messages over large distances in graphs and performs well on several graph classification benchmarks.
Cited by 1Open access
Jun Wang · 1991
Conditions of configuring feedforward neural networks without local minima are analyzed for both synchronous and asynchronous learning rules. Based on the analysis, a learning algorithm that integrates a synchronous-asynchronous learning rule with a dynamic configuration rule is presented to train feedforward neural networks. The theoretic analysis and numerical simulation reveal that the proposed learning algorithm substantially reduces the likelihood of local minimum solutions in supervised learning.
Cited by 0
Wong-bushby I, Passerini K, Patten Karen et al. · 2008
Discourse, a form of collaborative learning, is fundamentally a communications process. This in-progress study adapts Clark and Brennan’s grounding in communications principles to investigate how to “scaffold ” asynchronous discourse. Scaffolding is defined as providing support for the learner at his or her level until the support is no longer needed. This paper presents early results from an experimental study measuring learning effectiveness. In the experiment, content and process scaffolding are manipulated based on pedagogic principles. A major contribution of the study is building and testing a technologymediated, discourse-centered, teaching and learning model called the Asynchronous Learning Networks Cognitive Discourse Model (ALNCDM). As discourse is one of the most widely used online methods of teaching and learning, the results of the study are expected to add to the body of knowledge on how to structure asynchronous online discourse assignments for more effective student learning.
Cited by 0
Zhixiong Chen, Wenqiang Yi, Hyundong Shin et al. · IEEE Transactions on Communications · 2024
Owing to the heterogeneous computation and communication capabilities among clients, the synchronous model aggregation in wireless federated learning (FL) is susceptible to the straggler effect and exhibits low learning efficiency, while asynchronous aggregation encounters delayed gradients that lead to convergence errors and learning performance degradation. To address these obstacles, this work proposes an adaptive semi-asynchronous FL (ASAFL) approach to incorporate the strengths of synchronous and asynchronous FL while mitigating their inherent drawbacks. Specifically, the edge server dynamically adjusts the synchronous degree, i.e., the number of local gradients aggregated in each round, to strike a balance between learning latency and accuracy. Recognizing that data heterogeneity among clients may induce biased global model updating, we propose calibrating the global update by leveraging historical gradients received at the edge server from clients. Following that, we theoretically investigate the impact of synchronous degrees in different rounds on the convergence bound of ASAFL. The results imply that allocating more learning time to the later learning stages to increase the synchronous degree contributes to better learning performance. Based on this, we develop an adaptive synchronous degree control and resource allocation algorithm to enhance the learning performance of FL while adhering to the overall learning latency and wireless resources constraint. Numerical results on the MNIST and CIFAR-10 datasets demonstrate that the proposed approach is capable of attaining faster convergence speed and higher learning accuracy compared to the benchmark FL algorithms.
Cited by 20Open access
2004
Cited by 2
Dziuban, Charles, Shea, Peter, Arbaugh, J. B. · Journal of International Crisis and Risk Communication Research · 2004
Cited by 12Open access