Hasil untuk "History of scholarship and learning. The humanities"

David D. Zhang, Qing Pei, Shengda Zhang et al.

Abstract Outbreaks of epidemics are human ecological disasters and have caused huge losses of human life and social disturbances in human history. But their impact on human culture has never been systematically and quantitatively studied. This study hypothesizes that such gigantic human ecological pressure would have created a great need for cultural innovations. By quantitatively examining and modeling the process using the time-series of cultural innovations and human ecological–socioeconomic proxies in European history (1000–1900 CE) based on the basic principles of causal inference, the paper demonstrates that infectious disease epidemics and socioeconomic stress stimulated the flourishment of thinkers and philosophical thoughts across different philosophies in truth, knowledge, and ethics, and promoted scientific discovery/technological innovations in a macro scale. Based on the results of Poisson regression and analysis of marginal effects, when the epidemics increased by 1, the average number of philosophical thinkers increased by 0.85, and their average impact score increased by 4.04. When the Consumer Price Index (CPI) increased by 0.1, the average number of philosophical thinkers increased by 8.9, and the average impact score increased by 29.79. The results of linear regression further show that when the epidemics increased by 1, the average scientific discoveries and technological innovations (SDTI) increased by 0.128 units; when CPI increased by 10%, the average SDTI increased by 0.15 units. Infectious disease epidemics have generally played an important role in generating cultural dynamics during the study period. The results imply that the recurrent outbreaks of the COVID-19 pandemic would likely lead to another thriving phase of cultural innovations.

Li Cheng, Kimberly Phillips, Xiangting He

Given the need to guarantee sustainable, high-quality preschool education and considering preschool teachers’ central role in children’s learning and growth, the effective development of preschool teachers requires serious attention. This study explored the current intention to stay and the factors predicting retention intention among 214 urban preschool teachers in China using a retention intention scale, work value scale, and workload scale. The results showed that the preschool teachers’ retention intentions were above average, with age, marital status, major, years of experience, weekly working hours, salary, working relationship with the preschool, and the type and rank of the preschool significantly influencing this intent. Educational background, qualification certificate, and job title did not affect these intentions. The results indicated that workload was negatively associated with ECTs’ retention intentions, whereas work values were positively associated. Although the interaction between overall workload and total work values was not significant, further analysis of subdimensions revealed that intrinsic values amplified the negative effect of workload on retention intentions, while extrinsic values mitigated it. No significant moderating effect was found for external rewards. The findings highlight potential avenues for improving retention intention by focusing on intrinsic and extrinsic values and alleviating teacher workload in preschools and governmental institutions. This can, ultimately, have developmental benefits for both children and ECTs.

Faizah Hanim Zainuddin, Mashitah Hamidi, Abdul Muneem et al.

The objective of this study is to delve into the difficulties faced by older adults when it comes to receiving emotional support, and how these challenges affect their ability to age well. The study utilized a qualitative approach, conducting in-depth interviews and thematic analysis to present its findings. The sample group consisted of 24 older adults and three government officials who were interviewed. The study identified various obstacles to emotional support, such as family crises, stress from children’s problems, negative attitudes from children, and cultural changes. The analysis proposes several strategies to enhance emotional support for older adults, including encouraging them to volunteer to feel more connected, promoting the concept of self-care and self-love, utilizing technology to communicate, raising awareness about the emotional support needs of older adults, implementing intergenerational programs to promote emotional support across different age groups, and collaborating with community organizations to improve emotional support. Despite the importance of this topic, there is currently insufficient research on the barriers to emotional support in the context of successful aging, particularly in Malaysia.

Dao Cam Thuy, Nguyen Ngoc Quang

Selling through livestreaming has become a popular sales method and increasingly a preferred choice for retailers alongside traditional sales channels. Products that are intuitive and evoke purchasing emotions at first sight such as clothes, beauty products and household appliances are recorded with high revenue through livestream selling. This study aimed to evaluate the impact of various livestream sales elements including engaging sales content and sellers on customer attitude, trust and purchase intention for fashion products; investigate the factor of promotions moderating the relationship between the three factors above in the live shopping process. A linear structural model was used to analyze data from 324 customers shopping by livestream on TikTok. The results show that (1) Attractive content and streamers make significant contributions to the effectiveness of livestream sales. (2) engaging sales content, directly and indirectly, impacts purchasing intention through customer attitude and trust. (3) The influence of streamers does not have a direct impact but an indirect impact on purchase intention through customer attitude and trust. (4) Promotional programs influence purchase intention but do not moderate the relationship between trust and purchase intention and negatively moderate the relationship between attitude and purchase intention.

V. I. Kashcheev

The article deals with the problem of academic environment evolution in which an antiquity historian, A.S. Shofman (1913–1993) conducted his research and taught students in different periods of his life. The author shows that Shofman’s personal qualities, such as determination, ability to work in any external conditions, considerate and, at the same time, demanding attitude to his students and colleagues, sense of humor, irony, self-deprecation, and others played an important role in his development as a scholar. It is also pointed out that fate often brought him to teachers and colleagues who facilitated his success. Different spheres of his scientific communication are studied: prominent scholars of Ancient History and Philology – his teachers at Leningrad State University; international scientific contacts with his colleagues from Macedonia and Serbia, which used to be the part of Yugoslavia, from Hungary, Poland and Czechoslovakia. His relationship with Russian historians and philologists from Moscow, Leningrad, Voronezh, Tomsk and Belorussian scholars from Minsk were also discussed as well as academic environment at Kazan State University, especially at the Department of World History which A.S. Shofman headed for several decades. The article reveals the influence of his scientific surroundings on his development as a scholar and on themes of his research. A.S. Shofman, in his turn, shaped the mode of activity of the Department he headed and set up a scientific school in Kazan State University well-known in the country

Runxi Cheng, Yongxian Wei, Xianglong He et al.

Meta-learning has been extensively applied in the domains of few-shot learning and fast adaptation, achieving remarkable performance. While Meta-learning methods like Model-Agnostic Meta-Learning (MAML) and its variants provide a good set of initial parameters for the model, the model still tends to learn shortcut features, which leads to poor generalization. In this paper, we propose the formal conception of "learn to learn more precisely", which aims to make the model learn precise target knowledge from data and reduce the effect of noisy knowledge, such as background and noise. To achieve this target, we proposed a simple and effective meta-learning framework named Meta Self-Distillation(MSD) to maximize the consistency of learned knowledge, enhancing the models' ability to learn precise target knowledge. In the inner loop, MSD uses different augmented views of the same support data to update the model respectively. Then in the outer loop, MSD utilizes the same query data to optimize the consistency of learned knowledge, enhancing the model's ability to learn more precisely. Our experiment demonstrates that MSD exhibits remarkable performance in few-shot classification tasks in both standard and augmented scenarios, effectively boosting the accuracy and consistency of knowledge learned by the model.

Elena Martín-González, Ebraham Alskaf, Amedeo Chiribiri et al.

First-pass perfusion cardiac magnetic resonance (FPP-CMR) is becoming an essential non-invasive imaging method for detecting deficits of myocardial blood flow, allowing the assessment of coronary heart disease. Nevertheless, acquisitions suffer from relatively low spatial resolution and limited heart coverage. Compressed sensing (CS) methods have been proposed to accelerate FPP-CMR and achieve higher spatial resolution. However, the long reconstruction times have limited the widespread clinical use of CS in FPP-CMR. Deep learning techniques based on supervised learning have emerged as alternatives for speeding up reconstructions. However, these approaches require fully sampled data for training, which is not possible to obtain, particularly high-resolution FPP-CMR images. Here, we propose a physics-informed self-supervised deep learning FPP-CMR reconstruction approach for accelerating FPP-CMR scans and hence facilitate high spatial resolution imaging. The proposed method provides high-quality FPP-CMR images from 10x undersampled data without using fully sampled reference data.

Angela Zangaro

Manuel A. Castiñeiras González

Thomas Goerttler, Klaus Obermayer

In past years model-agnostic meta-learning (MAML) has been one of the most promising approaches in meta-learning. It can be applied to different kinds of problems, e.g., reinforcement learning, but also shows good results on few-shot learning tasks. Besides their tremendous success in these tasks, it has still not been fully revealed yet, why it works so well. Recent work proposes that MAML rather reuses features than rapidly learns. In this paper, we want to inspire a deeper understanding of this question by analyzing MAML's representation. We apply representation similarity analysis (RSA), a well-established method in neuroscience, to the few-shot learning instantiation of MAML. Although some part of our analysis supports their general results that feature reuse is predominant, we also reveal arguments against their conclusion. The similarity-increase of layers closer to the input layers arises from the learning task itself and not from the model. In addition, the representations after inner gradient steps make a broader change to the representation than the changes during meta-training.

Nahid Parween Anwar, Muhammad Abid Ali

Students should have the capability to argue about controversial science issues that are relevant to them and that impact society. These controversial issues, called socio-scientific issues (SSI), are influenced by social, ethical and moral norms. In current science education platforms, student-centred teaching strategies based on constructivism, are recommended to engage students in the construction of knowledge. Using a quantitative design, the present study sought to explore the efficacy of an argumentation-based teaching intervention about SSIs in an undergraduate classroom. It assessed students’ gains in the skill of argumentation. A one-group pre-test-post-test design was used. Data were generated through collection of students’ writing pre- and post-instruction, which was analysed using Toulmin’s Argumentation Pattern (TAP). Results reveal that almost half of the students reached a high level of argumentation. Findings show the importance of teaching content through learner-centred pedagogies. Introduction of various socio-scientific case studies and practicing argumentation has positively impacted on students’ argumentation skills. This study is significant for teachers as it provides an example to replicate in their classroom and can assist science teachers to enrich teaching and learning. The study recommends improving teachers’ competence in order to promote argumentation skills among students. Keywords: Argumentation in science, Socio-scientific Issue, Toulmin Argumentation Pattern, Genetics, Argumentation-based teaching How to cite this article: Anwar, N.P. 2020. The effect of socio-scientific issue (SSI) based discussion: A student-centred approach to the teaching of argumentation. Scholarship of Teaching and Learning in the South. 4(2): 35-62. https://doi.org/10.36615/sotls.v4i2.76. This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Brahma Pavse, Ishan Durugkar, Josiah Hanna et al.

Temporal difference (TD) learning is one of the main foundations of modern reinforcement learning. This paper studies the use of TD(0), a canonical TD algorithm, to estimate the value function of a given policy from a batch of data. In this batch setting, we show that TD(0) may converge to an inaccurate value function because the update following an action is weighted according to the number of times that action occurred in the batch -- not the true probability of the action under the given policy. To address this limitation, we introduce \textit{policy sampling error corrected}-TD(0) (PSEC-TD(0)). PSEC-TD(0) first estimates the empirical distribution of actions in each state in the batch and then uses importance sampling to correct for the mismatch between the empirical weighting and the correct weighting for updates following each action. We refine the concept of a certainty-equivalence estimate and argue that PSEC-TD(0) is a more data efficient estimator than TD(0) for a fixed batch of data. Finally, we conduct an empirical evaluation of PSEC-TD(0) on three batch value function learning tasks, with a hyperparameter sensitivity analysis, and show that PSEC-TD(0) produces value function estimates with lower mean squared error than TD(0).

Simon Vanneste, Astrid Vanneste, Kevin Mets et al.

Learning to communicate in order to share state information is an active problem in the area of multi-agent reinforcement learning (MARL). The credit assignment problem, the non-stationarity of the communication environment and the creation of influenceable agents are major challenges within this research field which need to be overcome in order to learn a valid communication protocol. This paper introduces the novel multi-agent counterfactual communication learning (MACC) method which adapts counterfactual reasoning in order to overcome the credit assignment problem for communicating agents. Secondly, the non-stationarity of the communication environment while learning the communication Q-function is overcome by creating the communication Q-function using the action policy of the other agents and the Q-function of the action environment. Additionally, a social loss function is introduced in order to create influenceable agents which is required to learn a valid communication protocol. Our experiments show that MACC is able to outperform the state-of-the-art baselines in four different scenarios in the Particle environment.

Alexandre Capone, Jonas Umlauft, Thomas Beckers et al.

The performance of learning-based control techniques crucially depends on how effectively the system is explored. While most exploration techniques aim to achieve a globally accurate model, such approaches are generally unsuited for systems with unbounded state spaces. Furthermore, a globally accurate model is not required to achieve good performance in many common control applications, e.g., local stabilization tasks. In this paper, we propose an active learning strategy for Gaussian process state space models that aims to obtain an accurate model on a bounded subset of the state-action space. Our approach aims to maximize the mutual information of the exploration trajectories with respect to a discretization of the region of interest. By employing model predictive control, the proposed technique integrates information collected during exploration and adaptively improves its exploration strategy. To enable computational tractability, we decouple the choice of most informative data points from the model predictive control optimization step. This yields two optimization problems that can be solved in parallel. We apply the proposed method to explore the state space of various dynamical systems and compare our approach to a commonly used entropy-based exploration strategy. In all experiments, our method yields a better model within the region of interest than the entropy-based method.

Chaochao Lu, Biwei Huang, Ke Wang et al.

Reinforcement learning (RL) algorithms usually require a substantial amount of interaction data and perform well only for specific tasks in a fixed environment. In some scenarios such as healthcare, however, usually only few records are available for each patient, and patients may show different responses to the same treatment, impeding the application of current RL algorithms to learn optimal policies. To address the issues of mechanism heterogeneity and related data scarcity, we propose a data-efficient RL algorithm that exploits structural causal models (SCMs) to model the state dynamics, which are estimated by leveraging both commonalities and differences across subjects. The learned SCM enables us to counterfactually reason what would have happened had another treatment been taken. It helps avoid real (possibly risky) exploration and mitigates the issue that limited experiences lead to biased policies. We propose counterfactual RL algorithms to learn both population-level and individual-level policies. We show that counterfactual outcomes are identifiable under mild conditions and that Q- learning on the counterfactual-based augmented data set converges to the optimal value function. Experimental results on synthetic and real-world data demonstrate the efficacy of the proposed approach.

Slavica Kozina, Martin Kowalski, Mirela Vlastelica et al.

Our aim was to determine (a) how parents deal with experiences like having a son missing in war, and (b) what expectations they have in terms of outcomes. This qualitative study included 29 parents of 21 sons gone missing in war. We used content analysis singling out narrative patterns and coded these. We assessed intercoder reliability using Krippendorff’s alpha coefficient. Items passing the Krippendorff’s alpha threshold of ≥.50 were verified using Cronbach’s alpha. Three of five coders showed acceptable intercoder agreement on 23 of the 173 identified topics (13.3%; Krippendorff’s alpha: .50-.82). Cronbach’s alpha coefficient confirmed intercoder reliability of .7903. Fathers’ narratives differ from mothers’. Statistics are a valuable tool for identifying specific motifs in grieving narratives of parents who have lost their child. Content analysis can provide insights without interfering with authentic personal experience sparing interviewees from reliving the traumatizing experience.

Pulkin Maxim Viktorovich

The article describes the basic laws of the penitentiary system’s formation in the Olonets province. The author identifies main problems that existed in the operation of prisons in the province as well as the ways of interaction of various departments in re-education and social adaptation of prisoners. The changes occurred in the prison system as results of the February Revolution of 1917 were studied.

Tri Dao, Albert Gu, Matthew Eichhorn et al.

Fast linear transforms are ubiquitous in machine learning, including the discrete Fourier transform, discrete cosine transform, and other structured transformations such as convolutions. All of these transforms can be represented by dense matrix-vector multiplication, yet each has a specialized and highly efficient (subquadratic) algorithm. We ask to what extent hand-crafting these algorithms and implementations is necessary, what structural priors they encode, and how much knowledge is required to automatically learn a fast algorithm for a provided structured transform. Motivated by a characterization of fast matrix-vector multiplication as products of sparse matrices, we introduce a parameterization of divide-and-conquer methods that is capable of representing a large class of transforms. This generic formulation can automatically learn an efficient algorithm for many important transforms; for example, it recovers the $O(N \log N)$ Cooley-Tukey FFT algorithm to machine precision, for dimensions $N$ up to $1024$. Furthermore, our method can be incorporated as a lightweight replacement of generic matrices in machine learning pipelines to learn efficient and compressible transformations. On a standard task of compressing a single hidden-layer network, our method exceeds the classification accuracy of unconstrained matrices on CIFAR-10 by 3.9 points -- the first time a structured approach has done so -- with 4X faster inference speed and 40X fewer parameters.

Jaehong Yoon, Saehoon Kim, Eunho Yang et al.

While recent continual learning methods largely alleviate the catastrophic problem on toy-sized datasets, some issues remain to be tackled to apply them to real-world problem domains. First, a continual learning model should effectively handle catastrophic forgetting and be efficient to train even with a large number of tasks. Secondly, it needs to tackle the problem of order-sensitivity, where the performance of the tasks largely varies based on the order of the task arrival sequence, as it may cause serious problems where fairness plays a critical role (e.g. medical diagnosis). To tackle these practical challenges, we propose a novel continual learning method that is scalable as well as order-robust, which instead of learning a completely shared set of weights, represents the parameters for each task as a sum of task-shared and sparse task-adaptive parameters. With our Additive Parameter Decomposition (APD), the task-adaptive parameters for earlier tasks remain mostly unaffected, where we update them only to reflect the changes made to the task-shared parameters. This decomposition of parameters effectively prevents catastrophic forgetting and order-sensitivity, while being computation- and memory-efficient. Further, we can achieve even better scalability with APD using hierarchical knowledge consolidation, which clusters the task-adaptive parameters to obtain hierarchically shared parameters. We validate our network with APD, APD-Net, on multiple benchmark datasets against state-of-the-art continual learning methods, which it largely outperforms in accuracy, scalability, and order-robustness.