Lifelong Learning Settings Research Articles

Increasing depth of neural networks for life-long learning

Purpose:We propose a novel method for continual learning based on the increasing depth of neural networks. This work explores whether extending neural network depth may be beneficial in a life-long learning setting. Methods:We propose a novel approach based on adding new layers on top of existing ones to enable the forward transfer of knowledge and adapting previously learned representations. We employ a method of determining the most similar tasks for selecting the best location in our network to add new nodes with trainable parameters. This approach allows for creating a tree-like model, where each node is a set of neural network parameters dedicated to a specific task. The Progressive Neural Network concept inspires the proposed method. Therefore, it benefits from dynamic changes in network structure. However, Progressive Neural Network allocates a lot of memory for the whole network structure during the learning process. The proposed method alleviates this by adding only part of a network for a new task and utilizing a subset of previously trained weights. At the same time, we may retain the benefit of PNN, such as no forgetting guaranteed by design, without needing a memory buffer. Results:Experiments on Split CIFAR and Split Tiny ImageNet show that the proposed algorithm is on par with other continual learning methods. In a more challenging setup with a single computer vision dataset as a separate task, our method outperforms Experience Replay. Conclusion:It is compatible with commonly used computer vision architectures and does not require a custom network structure. As an adaptation to changing data distribution is made by expanding the architecture, there is no need to utilize a rehearsal buffer. For this reason, our method could be used for sensitive applications where data privacy must be considered.

Exploring the associative learning capabilities of the segmented attractor network for lifelong learning.

This work explores the process of adapting the segmented attractor network to a lifelong learning setting. Taking inspirations from Hopfield networks and content-addressable memory, the segmented attractor network is a powerful tool for associative memory applications. The network's performance as an associative memory is analyzed using multiple metrics. In addition to the network's general hit rate, its capability to recall unique memories and their frequency is also evaluated with respect to time. Finally, additional learning techniques are implemented to enhance the network's recall capacity in the application of lifelong learning. These learning techniques are based on human cognitive functions such as memory consolidation, prediction, and forgetting.

Multilingual Machine Translation: Deep Analysis of Language-Specific Encoder-Decoders

State-of-the-art multilingual machine translation relies on a shared encoder-decoder. In this paper, we propose an alternative approach based on language-specific encoder-decoders, which can be easily extended to new languages by learning their corresponding modules. To establish a common interlingua representation, we simultaneously train N initial languages. Our experiments show that the proposed approach improves over the shared encoder-decoder for the initial languages and when adding new languages, without the need to retrain the remaining modules. All in all, our work closes the gap between shared and language-specific encoder-decoders, advancing toward modular multilingual machine translation systems that can be flexibly extended in lifelong learning settings.

The Extent to Which Ageism Has Affected Lifelong Learning Settings: Perspectives From Practitioners in Taiwan

As of 2018, Taiwan had a population of approximately 23.5 million people, of which 14% had reached 65 years old. To promote healthy and active aging for Taiwanese older adults, the Lifelong Learning Act was amended in 2018 in order to support older adults in terms of their personal development and leisure opportunities. The primary goal of the study was to explore the attitudes toward older adults and ageist behaviors shown by Taiwanese lifelong learning practitioners, including faculty members, lecturers, and volunteers. The results showed that Taiwanese lifelong learning practitioners, especially faculty members, had a high level of benevolent ageist attitudes toward older adults. However, associations between demographic factors on ageist behaviors were inconsistent. The findings of the study provide suggestions and research directions for the purpose of building a better age discrimination-free lifelong learning environment for current and prospective older learners.

Some Web-Based Experiences from Flipped Classroom Techniques in AEC Modules during the COVID-19 Lockdown

The classroom closure during the first semester of 2020 entailed decisive changes in higher education. Universities have become more digital in both the availability of e-resources and pervasive devices and how students communicate with lecturers and classmates. Learners adapted their study habits with a growing role of self-paced, internet-based strategies. Some flipped learning approaches have proven their efficacy under the remote-teaching physical constraints. This study aimed to appraise the outcomes from the implementation of various web-based, learning-aid tools on flipped teaching approaches in engineering modules. The open educational resources (OER) performed satisfactorily during the lockdown period in three universities from two countries with similar higher education models. Such resources encompassed classroom response systems and web-based exercise repositories, designed for diverse purposes such as autonomous learning, self-correction, flipped classroom, peer assessment, and guided study. The acquired experiences reveal that OER helped students to enhance their engagement, reach the deeper levels of the cone of learning, and widen their range of learning abilities. This procedure is easily attainable for architecture, engineering, and construction (AEC) courses and lifelong learning settings. Feedback from students, instructors’ perceptions, and learning outcomes show the suitability and effectiveness of the web-based learning assistant procedure presented here.

Lifelong Visual-Tactile Cross-Modal Learning for Robotic Material Perception.

The material attribute of an object's surface is critical to enable robots to perform dexterous manipulations or actively interact with their surrounding objects. Tactile sensing has shown great advantages in capturing material properties of an object's surface. However, the conventional classification method based on tactile information may not be suitable to estimate or infer material properties, particularly during interacting with unfamiliar objects in unstructured environments. Moreover, it is difficult to intuitively obtain material properties from tactile data as the tactile signals about material properties are typically dynamic time sequences. In this article, a visual-tactile cross-modal learning framework is proposed for robotic material perception. In particular, we address visual-tactile cross-modal learning in the lifelong learning setting, which is beneficial to incrementally improve the ability of robotic cross-modal material perception. To this end, we proposed a novel lifelong cross-modal learning model. Experimental results on the three publicly available data sets demonstrate the effectiveness of the proposed method.

„Gesundheitsförderung – lohnt sich das?“ Ein Diskussionsspiel für Ausbildung und Praxis

In order to strengthen health promotion spreading knowledge is as essential as tackling typical problems of its implementation. Both aspects can be adressed by using game-based learning approaches. To this end, we specifically developed a discussion game divided into four phases of play. Our aim is to (1) spread facts, (2) raise awareness of the topic and (3) promote well-informed political opinion-forming at individual as well as group levels. Experience shows that these objectives can be achieved with this game. The game as a whole or parts of it can easily be integrated into university teaching in different ways. In addition, there is major potential for using the game in lifelong learning settings and health promotion projects. In the next step the game will be adapted to enable its implementation in the workplace as a special setting for health promotion.

Using Task Descriptions in Lifelong Machine Learning for Improved Performance and Zero-Shot Transfer

Knowledge transfer between tasks can improve the performance of learned models, but requires an accurate estimate of inter-task relationships to identify the relevant knowledge to transfer. These inter-task relationships are typically estimated based on training data for each task, which is inefficient in lifelong learning settings where the goal is to learn each consecutive task rapidly from as little data as possible. To reduce this burden, we develop a lifelong learning method based on coupled dictionary learning that utilizes high-level task descriptions to model inter-task relationships. We show that using task descriptors improves the performance of the learned task policies, providing both theoretical justification for the benefit and empirical demonstration of the improvement across a variety of learning problems. Given only the descriptor for a new task, the lifelong learner is also able to accurately predict a model for the new task through zero-shot learning using the coupled dictionary, eliminating the need to gather training data before addressing the task.

Toward Training Recurrent Neural Networks for Lifelong Learning.

Catastrophic forgetting and capacity saturation are the central challenges of any parametric lifelong learning system. In this work, we study these challenges in the context of sequential supervised learning with an emphasis on recurrent neural networks. To evaluate the models in the lifelong learning setting, we propose a curriculum-based, simple, and intuitive benchmark where the models are trained on tasks with increasing levels of difficulty. To measure the impact of catastrophic forgetting, the model is tested on all the previous tasks as it completes any task. As a step toward developing true lifelong learning systems, we unify gradient episodic memory (a catastrophic forgetting alleviation approach) and Net2Net (a capacity expansion approach). Both models are proposed in the context of feedforward networks, and we evaluate the feasibility of using them for recurrent networks. Evaluation on the proposed benchmark shows that the unified model is more suitable than the constituent models for lifelong learning setting.

An empirical investigation of the antecedents of learner-centered outcome measures in MOOCs

This research revealed the antecedes of two learner-centered outcome measures of success in massive open online courses (MOOCs): learner satisfaction and learner intention-fulfillment. Previous studies used success criteria from formal education contexts placing retention and completion rates as the ultimate outcome measures. We argue that the suggested learner-centered outcomes are more appropriate for measuring success in non-formal lifelong learning settings because they are focused on the learner’s intentions, rather than the intentions of the course developer. The behavioural measures of 125 MOOC participants who answered a pre- and a post-questionnaire were harvested. The analysis revealed that learner satisfaction was directly affected by: the importance of the MOOC’s benefits; online self-regulated learning - goal setting; number of video lectures accessed; and, perceived course usability. Age and the number of quizzes accessed indirectly effected learner satisfaction, through perceived course usability and through number of video lectures accessed. Intention-fulfillment was directly affected by: gender; the importance of the MOOC’s benefits; online self-regulated learning - goal setting; the number of quizzes accessed; the duration of participation; and, perceived course usability. Previous experience with MOOCs and the importance of MOOC’s benefits, indirectly affected intention-fulfillment through the number of quizzes accessed and perceived course usability.

Presenting the outline of a proposal for a 5 part program of medical education research using eLearning or Technology enhanced learning to support Learning through the continuum of Undergraduate, through Postgraduate to Lifelong learning settings

This article was migrated. The article was not marked as recommended. The purpose of this article is to present the outline of a proposal for a 5 part program of medical education (MedEd) research using eLearning or Technology enhanced learning (TeL) to support Learning through the continuum of Undergraduate, through Postgraduate to Lifelong learning settings.

Disaster preparedness, adaptive politics and lifelong learning: a case of Japan

Preparedness for disaster scenarios is progressively becoming an educational agenda for governments because of diversifying risks and threats worldwide. In disaster-prone Japan, disaster preparedness has been a prioritised national agenda, and preparedness education has been undertaken in both formal schooling and lifelong learning settings. This article examines the politics behind one prevailing policy discourse in the field of disaster preparedness referred to as ‘the four forms of aid’ – ‘kojo [public aid]’, ‘jijo [self-help]’, ‘gojo/kyojo [mutual aid]’. The study looks at the Japanese case, however, the significance is global, given that neo-liberal governments are increasingly having to deal with a range of disaster situations whether floods or terrorism, while implementing austerity measures. Drawing on the theory of the adaptiveness of neo-liberalism, the article sheds light on the hybridity of the current Abe government’s politics: a ‘dominant’ neo-liberal economic approach – public aid and self-help – and a ‘subordinate’ moral conservative agenda – mutual aid. It is argued that the four forms of aid are an effective ‘balancing act’, and that kyojo in particular is a powerful legitimator in the hybrid politics. The article concludes that a lifelong and life-wide preparedness model could be developed in Japan which has taken a social approach to lifelong learning.

Universities as spaces for engaging the other: A pedagogy of encounter for intercultural and interreligious education

Internationalisation of Higher Education coupled with students’ mobility across the globe in the 21st century has led to universities in many countries having to deal with an unprecedented flow of human diversity. Such diversity can become a potential source of conflict due to increased “otherness”, but at the same time, it also presents an ideal lifelong learning opportunity for engaging the other through intercultural and interreligious education. This paper begins with a literature review, exploring the challenges of international higher education and the opportunities it offers in fostering intercultural and interreligious dialogue and deeper engagement across values, beliefs, world views, habits of mind and ways of being. The review highlights that on the one hand, rich diversity and otherness within universities can easily be built up in so-called educated individuals as givens and therefore be treated with insensitivity or indifference. On the other hand, diversity can evoke the possibility for the exchange of thoughts, values and world views, sharing experiences, engaging with each other’s foreignness, and making the encounter a real enrichment and transformation. Next, using the narrative of a case study about conflicts within an international students’ residence, the author illustrates the necessity of encounter and engagement with otherness as decisive avenues for intercultural learning and interreligious understanding. Through a further analysis of the case study, she establishes that pedagogical strategies formulated around encounter, dialogue and engagement should be integrated into the students’ life at international universities so that they serve to bridge religions, cultures, world views and other differences, thus creating a sustainable culture of dialogue and peace. The paper concludes by suggesting a few key elements which seem useful for implementing a pedagogy of encounter in both formal and informal lifelong learning settings.

Analysing entrepreneurship education: a bibliometric survey pattern

Entrepreneurship education is an evolving field that confronts obstacles due to fragmentation issues and eclectic approaches that have to be resolved utilising robust educational theories and tools able to intrude effectively the entrepreneurial research discourse. Entrepreneurial learning is also the outcome of education and an unequivocal component of theorising about entrepreneurship. Based on explanatory bibliometric techniques, the present study examines, for the first time, how these terms have emerged in the extant entrepreneurship literature since eighties. A set of 7726 abstracts, retrieved from the SCOPUS database, is analysed through (key)word frequencies, co-occurrence networks and citations. Quantitative findings verify the customary picture for entrepreneurship education that exhibits low academic citation and loose connections with learning theories. The present data also reveal that the connection of entrepreneurship with lifelong learning settings, vocational training and career counselling is scarce in literature. Other ‘gaps’ in research pertain to the comprehensive examination of experiential learning, advanced learning processes and education for innovation. The quantitatively identified shortage of the previous research topics is crucial for the future development of the field of entrepreneurship. Implications concern educational researchers in the field of entrepreneurship, educational agencies or policies as well as academic publishers.

Online Multi-Task Learning via Sparse Dictionary Optimization

This paper develops an efficient online algorithm for learning multiple consecutive tasks based on the K-SVD algorithm for sparse dictionary optimization. We first derive a batch multi-task learning method that builds upon K-SVD, and then extend the batch algorithm to train models online in a lifelong learning setting. The resulting method has lower computational complexity than other current lifelong learning algorithms while maintaining nearly identical model performance. Additionally, the proposed method offers an alternate formulation for lifelong learning that supports both task and feature similarity matrices.

Active Task Selection for Lifelong Machine Learning

In a lifelong learning framework, an agent acquires knowledge incrementally over consecutive learning tasks, continually building upon its experience. Recent lifelong learning algorithms have achieved nearly identical performance to batch multi-task learning methods while reducing learning time by three orders of magnitude. In this paper, we further improve the scalability of lifelong learning by developing curriculum selection methods that enable an agent to actively select the next task to learn in order to maximize performance on future learning tasks. We demonstrate that active task selection is highly reliable and effective, allowing an agent to learn high performance models using up to 50% fewer tasks than when the agent has no control over the task order. We also explore a variant of transfer learning in the lifelong learning setting in which the agent can focus knowledge acquisition toward a particular target task.

Educational patterns in lifelong learning settings

(1988). Educational patterns in lifelong learning settings. International Journal of Lifelong Education: Vol. 7, No. 3, pp. 205-221.