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

Enrique Flores

La vorágine es una gran obra disruptiva en el horizonte de la narrativa latinoamericana. Su radical desafío psíquico y su vínculo con lo fantasmal la vuelven única entre las novelas de la selva. Este ensayo profundiza su dimensión “espectral” asociada a la fotografía —aunque no únicamente a ella— a partir de otros estudios, del marco teórico expuesto en La cámara lúcida por Roland Barthes, del lugar de la fotografía en la novela, y de la figura de Eugène Robuchon, el fotógrafo francés cuya experiencia sirvió de modelo a José Eustasio Rivera.

Kai Bao, Meihua Liu

This study compared the five-word lexical bundles (LBs) expressing gratitude in acknowledgments of dissertations written by Chinese and American PhD students of linguistics. Two corpora were built: (1) The Chinese University Dissertation Acknowledgments Collection (CUC) which contained 700 acknowledgments with a total of 300,686 tokens, and (2) the American University Dissertation Acknowledgments Collection (AUC) which contained 700 acknowledgments with a total of 493,045 tokens. We then retrieved five-word LBs, of which LBs expressing gratitude in CUC and AUC were identified, categorized, and compared with respect to frequency, forms and structures. Major findings were: (1) the Chinese students used a substantially greater number of gratitude LBs than the American students, (2) the two groups used considerably different gratitude LBs, and (3) the two groups mainly relied on verb phrase-based LBs to express gratitude, but the Chinese students used a larger proportion of noun phrase- yet a smaller proportion of verb phrase-based items than the American students, and (4) the two groups used dissimilar structures and words to construct gratitude LBs. These findings enrich our knowledge of linguistic patterns in dissertation acknowledgments as a unique genre of academic prose, and provide corpus-based learning materials for students tasked with properly expressing gratitude in their theses or dissertations.

Čorboloković Saša S., Gavranović Valentina M.

The paper investigates punctuation rules and their application in Serbian and English, focusing on the examples that comply with different normative solutions in the two languages. The main goal of the research is to compare and contrast the results obtained from a survey done by a group of seventh-grade primary school students. The paper aims to determine how well the respondents apply punctuation rules in Serbian and English, to examine whether there is interference in the application of rules, and to investigate to what extent the detected errors illustrate the tendency of spreading pseudo-norms that violate the orthography of both languages. The results show that the respondents use punctuation marks with more precision in Serbian than in English. The percentage of incorrect answers to each question and the types of errors indicate interference and the creation of hybrid forms that are incorrect in both languages, which represent the creation of pseudo-norms. Furthermore, the results show a greater influence of the application of the rules adopted in the Serbian language on the English language, which can be interpreted by the bigger number of Serbian classes and clearly stated topics within the syllabus of the Serbian language course.

Alireza Mousavi-Hosseini, Adel Javanmard, Murat A. Erdogdu

Recently, there have been numerous studies on feature learning with neural networks, specifically on learning single- and multi-index models where the target is a function of a low-dimensional projection of the input. Prior works have shown that in high dimensions, the majority of the compute and data resources are spent on recovering the low-dimensional projection; once this subspace is recovered, the remainder of the target can be learned independently of the ambient dimension. However, implications of feature learning in adversarial settings remain unexplored. In this work, we take the first steps towards understanding adversarially robust feature learning with neural networks. Specifically, we prove that the hidden directions of a multi-index model offer a Bayes optimal low-dimensional projection for robustness against $\ell_2$-bounded adversarial perturbations under the squared loss, assuming that the multi-index coordinates are statistically independent from the rest of the coordinates. Therefore, robust learning can be achieved by first performing standard feature learning, then robustly tuning a linear readout layer on top of the standard representations. In particular, we show that adversarially robust learning is just as easy as standard learning. Specifically, the additional number of samples needed to robustly learn multi-index models when compared to standard learning does not depend on dimensionality.

Andrzej Samborski

W pracy wykorzystano dane ze stacji meteorologicznych funkcjonujących na terenie Zamościa w okresie od 1976 do 2020 roku. Opisano zmienność wartości wybranych elementów meteorologicznych i charakterystyk klimatycznych w skali czasowej, wykorzystując najczęściej stosowane statystyki rozkładu, tzn. wartości średnie oraz miary zmienności. Wyznaczono kierunek trendu zmian temperatury powietrza. W celu określenia warunków bioklimatycznych obliczono wartości wybranych wskaźników, takich jak: temperatura ekwiwalentna, temperatura efektywna i wielkość ochładzająca powietrza. Przeprowadzone badania wskazują, że w Zamościu optymalne warunki termiczne określane jako przyjemnie chłodno występują w okresie letnim (czerwiec, lipiec, sierpień), przy czym w lipcu są one łagodne. W kwietniu, maju i we wrześniu odczucie cieplne charakteryzowane jest jako chłodne. Generalnie w okresie od listopada do marca jest zimno, a w styczniu i w lutym bardzo zimno.

Yenchun Jim Wu, Mark Goh, Yingping Mai

Abstract Scholarly attention to social innovation has increased rapidly in recent years, but a broad picture to illustrate the structure of the field, tracing its evolution and identifying new research areas has been lacking. Thus, we conduct a three-step method to explore the trajectory of academic research on social innovation and identify potential research opportunities. In the first step, the bibliometric mapping software VOSviewer is used to visualize the network of authors and keyword co-occurrences. Next, SciMAT is applied to illustrate the evolution and importance of the themes. Then, content analysis is used to identify how specific research topics in social innovation in higher education have evolved. The author-keyword occurrence analysis and evolution map results reveal that innovation and design thinking were the most prominent keywords, and the citation analysis indicates that the works of Swyngedouw, Moulaert, and Westley as the leading research works in this field. The most influential countries during the study period from 1996 to July 2021 were the UK, the US, and Italy. Furthermore, the research collaboration network is more active among developed countries than among developing countries. Three research streams in social innovation in higher education have been identified: curriculum transformation, community-university partnership, and helix partnerships. This study supports understanding how higher education is shaped through social innovation. The results from this study can serve as a research reference on the state of the art and point to new research opportunities, notably on the need to engage developing countries in collaborating on social innovation in higher education through design, engagement, and partnership.

Asmaa Al-suhimi

The current study aimed to design a counseling program based on gratitude for the development of psychological well-being, verify the effectiveness of this program and identify the continuity of its impact through the results of follow-up measurement after a month of its application. An intentional sample of (24) female students from the College of Education at King Khalid University, were randomly divided into two groups, a control, and an experimental group, each group consisted of 12 students, and the research tools used to measure the psychological well-being of university students Prepared by Shend et al. (2013), gratitude scale prepared by Arnout (2019b), and a program based on gratitude consisting of 11 sessions prepared by the researcher. The results found that there were statistically significant differences at the level (0.05) between the means of the control and experimental groups in gratitude and psychological well-being in the post-measurement in favor of the experimental group, and also there were statistically significant differences at the level (0.05) between the means of the experimental group in the pre and post- measurement in gratitude and psychological well-being in favor of the post-measurement, and this indicates the effectiveness of the program. The results also revealed that there were no statistically significant differences between the mean scores of the experimental group in the post and follow-up measurements of gratitude and psychological well-being, which indicates the continuity of improvement and the effect of the counseling program based on gratitude for the development of psychological well-being.

Pranjal Awasthi, Nika Haghtalab, Eric Zhao

Multi-distribution learning is a natural generalization of PAC learning to settings with multiple data distributions. There remains a significant gap between the known upper and lower bounds for PAC-learnable classes. In particular, though we understand the sample complexity of learning a VC dimension d class on $k$ distributions to be $O(ε^{-2} \ln(k)(d + k) + \min\{ε^{-1} dk, ε^{-4} \ln(k) d\})$, the best lower bound is $Ω(ε^{-2}(d + k \ln(k)))$. We discuss recent progress on this problem and some hurdles that are fundamental to the use of game dynamics in statistical learning.

Quan Nguyen, Nishant A. Mehta

We consider the adversarial online multi-task reinforcement learning setting, where in each of $K$ episodes the learner is given an unknown task taken from a finite set of $M$ unknown finite-horizon MDP models. The learner's objective is to minimize its regret with respect to the optimal policy for each task. We assume the MDPs in $\mathcal{M}$ are well-separated under a notion of $λ$-separability, and show that this notion generalizes many task-separability notions from previous works. We prove a minimax lower bound of $Ω(K\sqrt{DSAH})$ on the regret of any learning algorithm and an instance-specific lower bound of $Ω(\frac{K}{λ^2})$ in sample complexity for a class of uniformly-good cluster-then-learn algorithms. We use a novel construction called 2-JAO MDP for proving the instance-specific lower bound. The lower bounds are complemented with a polynomial time algorithm that obtains $\tilde{O}(\frac{K}{λ^2})$ sample complexity guarantee for the clustering phase and $\tilde{O}(\sqrt{MK})$ regret guarantee for the learning phase, indicating that the dependency on $K$ and $\frac{1}{λ^2}$ is tight.

Nasraddin Mohammed

ئامانجى ئه‌م توێژینه‌وه‌یه‌ زانینى ئاراسته‌ى سه‌مپلێك له‌ قوتابیانى زانكۆى زاخۆیه‌ به‌رامبه‌ر به‌كوردیكردنى پرۆگرامى خوێندن له‌ زانكۆكانى هه‌رێمى كوردستان و په‌یوه‌ندى به‌ گۆڕاوه‌كانى ره‌گه‌ز، پسپۆرى، ته‌مه‌ن و قۆناغى خوێندن. سه‌مپلى توێژینه‌وه‌ پێكهاتووه‌ له‌ (235) قوتابى زانكۆى ناوبراو كه‌ به‌ شێوه‌یه‌كى هه‌ڕمه‌كى هه‌ڵبژێردراون، و بۆ ئه‌م مه‌به‌سته‌ توێژه‌ر هه‌ڵساوه‌ به‌ ئاماده‌كردنى راپرسییه‌ك كه‌ (32) بڕگه‌ له‌خۆ ده‌گرێت، و دواى ده‌رهێنانى تایبه‌تمه‌ندییه‌كانى سایكۆمه‌تریك، و بۆ چاره‌سه‌رى داتاكان ئامرازى ئامارى گۆنجاوى به‌كارهێناوه‌. و ئه‌نجامه‌كان پیشانیانداوه‌ ئاستێكى ئه‌رێنى له‌ ئاراسته‌ى قوتابیانى زانكۆى زاخۆ به‌رامبه‌ر به‌كوردیكردنى پرۆگرامى خوێندن له‌و قۆناغه‌ به‌دى ده‌كرێت، و له‌گه‌ڵ ئه‌وه‌شدا هیچ جیاوازییه‌كى واتادارى ئامارى له‌ نێوان ئاراسته‌كانیان به‌ پێى گۆڕاوه‌كانى توێژینه‌وه‌ نه‌بوو، و له‌سایه‌ى ئه‌م ئه‌نجامانه‌ توێژه‌ر كۆمه‌ڵێك ڕاسپارده‌ و پێشنیارى خسته‌روو.

Zhishuai Guo, Rong Jin, Jiebo Luo et al.

In this paper, we tackle a novel federated learning (FL) problem for optimizing a family of X-risks, to which no existing FL algorithms are applicable. In particular, the objective has the form of $\mathbb E_{z\sim S_1} f(\mathbb E_{z'\sim S_2} \ell(w; z, z'))$, where two sets of data $S_1, S_2$ are distributed over multiple machines, $\ell(\cdot)$ is a pairwise loss that only depends on the prediction outputs of the input data pairs $(z, z')$, and $f(\cdot)$ is possibly a non-linear non-convex function. This problem has important applications in machine learning, e.g., AUROC maximization with a pairwise loss, and partial AUROC maximization with a compositional loss. The challenges for designing an FL algorithm for X-risks lie in the non-decomposability of the objective over multiple machines and the interdependency between different machines. To this end, we propose an active-passive decomposition framework that decouples the gradient's components with two types, namely active parts and passive parts, where the active parts depend on local data that are computed with the local model and the passive parts depend on other machines that are communicated/computed based on historical models and samples. Under this framework, we develop two provable FL algorithms (FeDXL) for handling linear and nonlinear $f$, respectively, based on federated averaging and merging. We develop a novel theoretical analysis to combat the latency of the passive parts and the interdependency between the local model parameters and the involved data for computing local gradient estimators. We establish both iteration and communication complexities and show that using the historical samples and models for computing the passive parts do not degrade the complexities. We conduct empirical studies of FeDXL for deep AUROC and partial AUROC maximization, and demonstrate their performance compared with several baselines.

Davide Piras, Hiranya V. Peiris, Andrew Pontzen et al.

We develop the use of mutual information (MI), a well-established metric in information theory, to interpret the inner workings of deep learning models. To accurately estimate MI from a finite number of samples, we present GMM-MI (pronounced $``$Jimmie$"$), an algorithm based on Gaussian mixture models that can be applied to both discrete and continuous settings. GMM-MI is computationally efficient, robust to the choice of hyperparameters and provides the uncertainty on the MI estimate due to the finite sample size. We extensively validate GMM-MI on toy data for which the ground truth MI is known, comparing its performance against established mutual information estimators. We then demonstrate the use of our MI estimator in the context of representation learning, working with synthetic data and physical datasets describing highly non-linear processes. We train deep learning models to encode high-dimensional data within a meaningful compressed (latent) representation, and use GMM-MI to quantify both the level of disentanglement between the latent variables, and their association with relevant physical quantities, thus unlocking the interpretability of the latent representation. We make GMM-MI publicly available.

Jianfei Yang, Xinyan Chen, Dazhuo Wang et al.

WiFi sensing has been evolving rapidly in recent years. Empowered by propagation models and deep learning methods, many challenging applications are realized such as WiFi-based human activity recognition and gesture recognition. However, in contrast to deep learning for visual recognition and natural language processing, no sufficiently comprehensive public benchmark exists. In this paper, we review the recent progress on deep learning enabled WiFi sensing, and then propose a benchmark, SenseFi, to study the effectiveness of various deep learning models for WiFi sensing. These advanced models are compared in terms of distinct sensing tasks, WiFi platforms, recognition accuracy, model size, computational complexity, feature transferability, and adaptability of unsupervised learning. It is also regarded as a tutorial for deep learning based WiFi sensing, starting from CSI hardware platform to sensing algorithms. The extensive experiments provide us with experiences in deep model design, learning strategy skills and training techniques for real-world applications. To the best of our knowledge, this is the first benchmark with an open-source library for deep learning in WiFi sensing research. The benchmark codes are available at https://github.com/xyanchen/WiFi-CSI-Sensing-Benchmark.

Eseoghene Ben-Iwhiwhu, Saptarshi Nath, Praveen K. Pilly et al.

Lifelong learning aims to create AI systems that continuously and incrementally learn during a lifetime, similar to biological learning. Attempts so far have met problems, including catastrophic forgetting, interference among tasks, and the inability to exploit previous knowledge. While considerable research has focused on learning multiple supervised classification tasks that involve changes in the input distribution, lifelong reinforcement learning (LRL) must deal with variations in the state and transition distributions, and in the reward functions. Modulating masks with a fixed backbone network, recently developed for classification, are particularly suitable to deal with such a large spectrum of task variations. In this paper, we adapted modulating masks to work with deep LRL, specifically PPO and IMPALA agents. The comparison with LRL baselines in both discrete and continuous RL tasks shows superior performance. We further investigated the use of a linear combination of previously learned masks to exploit previous knowledge when learning new tasks: not only is learning faster, the algorithm solves tasks that we could not otherwise solve from scratch due to extremely sparse rewards. The results suggest that RL with modulating masks is a promising approach to lifelong learning, to the composition of knowledge to learn increasingly complex tasks, and to knowledge reuse for efficient and faster learning.

Prateek Gupta, Elias B. Khalil, Didier Chetélat et al.

The expressive and computationally inexpensive bipartite Graph Neural Networks (GNN) have been shown to be an important component of deep learning based Mixed-Integer Linear Program (MILP) solvers. Recent works have demonstrated the effectiveness of such GNNs in replacing the branching (variable selection) heuristic in branch-and-bound (B&B) solvers. These GNNs are trained, offline and on a collection of MILPs, to imitate a very good but computationally expensive branching heuristic, strong branching. Given that B&B results in a tree of sub-MILPs, we ask (a) whether there are strong dependencies exhibited by the target heuristic among the neighboring nodes of the B&B tree, and (b) if so, whether we can incorporate them in our training procedure. Specifically, we find that with the strong branching heuristic, a child node's best choice was often the parent's second-best choice. We call this the "lookback" phenomenon. Surprisingly, the typical branching GNN of Gasse et al. (2019) often misses this simple "answer". To imitate the target behavior more closely by incorporating the lookback phenomenon in GNNs, we propose two methods: (a) target smoothing for the standard cross-entropy loss function, and (b) adding a Parent-as-Target (PAT) Lookback regularizer term. Finally, we propose a model selection framework to incorporate harder-to-formulate objectives such as solving time in the final models. Through extensive experimentation on standard benchmark instances, we show that our proposal results in up to 22% decrease in the size of the B&B tree and up to 15% improvement in the solving times.

Javed Aslam, Aqeela Saleem, Nokhaiz Tariq Khan et al.

Planning to adopt the Blockchain is very active in many industries, especially in supply chains. Researchers believe that the Radio-frequency identification (RFIDs), yesterday’s Blockchain, is now obsolete. The strongest reason that the Blockchain is the tool of this era is its unique features; real-time information sharing, cyber-security, transparency, reliability, traceability, and visibility, all of which boost the supply-chain performance. Despite the extensive literature on Blockchain, in recent years, no clear framework has defined whether a supply chain should implement Blockchain or not. This study attempts to fill this gap by proposing a framework for complex supply chain networks. In doing so, first, we identified the supply-chain practices of the oil industry in Pakistan, then we empirically analyzed the impact of these practices on operational performance. The results show that the supply chain management (SCM) practices positively impact operational performance. On the other hand, with the help of literature, we identified different Blockchain features and their influence on different supply chain practices. This study guides managers and decision-makers to evaluate their current supply-chain practices and understand the relationship between supply-chain practices and Blockchain features, and how different Blockchain features can help improving supply-chain practices and ultimately improving operational performance.

Yi Liu, Congping Li, Yuan Li

Abstract Traditional studies on urban pathology primarily focused on impacts of social disorganization and urban settings, such as crime, polarization between the rich and the poor, pollution, and deteriorating living conditions, whereas there is insufficient attention paid to urban mental health and wellness. To provide fresh insights into the issue of urban mental health and wellness, this paper defines the psychiatric disorder of urban citizens as ‘new urban pathology’ which primarily results from a highly competitive, stressful, and fast-paced urban life. Based on 40 interviews from citizens in two rapidly urbanizing cities in mainland China, this paper attempts to investigate how the supply of leisure facilities affects new urban pathology. There are three general findings of this research: (1) First, new urban pathology commonly exists and varies with urban communities. However, it is not significantly influenced by the pace of city life. It is found that the worse prospect the living standards, the more significant the negative impacts of the new urban pathology. (2) Sufficient urban leisure facilities do have positive impacts on relieving psychological pressure of urban dwellers. Apart from these, residents also achieve stress relief from surrounding therapeutic landscapes like aesthetic public space, well-designed architecture, etc. (3) The perceived usefulness of leisure environment serves as a key factor to mediate the stress-mitigating effect of leisure supply. When leisure facilities’ functions highly match the needs of residents, leisure environmental supply can be utilized effectively, which helps alleviate the new urban pathology. This research advances the literature of urban health studies by tentatively revealing the interaction between the supply of leisure environmental facilities and urban mental health. It suggests that urban practitioners should optimize the quality of facilities rather than simply increasing the quantity for reducing the daily stress of urban life.

Dobrosława Wężowicz-Ziółkowska

Культурные источники антропопрессии [Рецензия на книгу Różnice kulturowe w traktowaniu zwierząt. Ред. Hanna Mamzer. Oficyna Wydawnicza ATUT, Wrocław 2020]

Matias Valdenegro-Toro

Uncertainty in machine learning is not generally taught as general knowledge in Machine Learning course curricula. In this paper we propose a short curriculum for a course about uncertainty in machine learning, and complement the course with a selection of use cases, aimed to trigger discussion and let students play with the concepts of uncertainty in a programming setting. Our use cases cover the concept of output uncertainty, Bayesian neural networks and weight distributions, sources of uncertainty, and out of distribution detection. We expect that this curriculum and set of use cases motivates the community to adopt these important concepts into courses for safety in AI.

Tanya Ignatenko, Kirill Kondrashov, Marco Cox et al.

User preference learning is generally a hard problem. Individual preferences are typically unknown even to users themselves, while the space of choices is infinite. Here we study user preference learning from information-theoretic perspective. We model preference learning as a system with two interacting sub-systems, one representing a user with his/her preferences and another one representing an agent that has to learn these preferences. The user with his/her behaviour is modeled by a parametric preference function. To efficiently learn the preferences and reduce search space quickly, we propose the agent that interacts with the user to collect the most informative data for learning. The agent presents two proposals to the user for evaluation, and the user rates them based on his/her preference function. We show that the optimum agent strategy for data collection and preference learning is a result of maximin optimization of the normalized weighted Kullback-Leibler (KL) divergence between true and agent-assigned predictive user response distributions. The resulting value of KL-divergence, which we also call remaining system uncertainty (RSU), provides an efficient performance metric in the absence of the ground truth. This metric characterises how well the agent can predict user and, thus, the quality of the underlying learned user (preference) model. Our proposed agent comprises sequential mechanisms for user model inference and proposal generation. To infer the user model (preference function), Bayesian approximate inference is used in the agent. The data collection strategy is to generate proposals, responses to which help resolving uncertainty associated with prediction of the user responses the most. The efficiency of our approach is validated by numerical simulations. Also a real-life example of preference learning application is provided.