Search

Low-resolution coarse-grained (CG) models provide significant computational and conceptual advantages for simulating soft materials. However, the properties of CG models depend quite sensitively upon the mapping, M, that maps each atomic configuration, r, to a CG configuration, R. In particular, M determines how the configurational information of the atomic model is partitioned between the mapped ensemble of CG configurations and the lost ensemble of atomic configurations that map to each R. In this work, we investigate how the mapping partitions the atomic configuration space into CG and intra-site components. We demonstrate that the corresponding coordinate transformation introduces a nontrivial Jacobian factor. This Jacobian factor defines a labeling entropy that corresponds to the uncertainty in the atoms that are associated with each CG site. Consequently, the labeling entropy effectively transfers configurational information from the lost ensemble into the mapped ensemble. Moreover, our analysis highlights the possibility of resonant mappings that separate the atomic potential into CG and intra-site contributions. We numerically illustrate these considerations with a Gaussian network model for the equilibrium fluctuations of actin. We demonstrate that the spectral quality, Q , provides a simple metric for identifying high quality representations for actin. Conversely, we find that neither maximizing nor minimizing the information content of the mapped ensemble results in high quality representations. However, if one accounts for the labeling uncertainty, Q (M) correlates quite well with the adjusted configurational information loss, I ̂ map (M) , that results from the mapping. [ABSTRACT FROM AUTHOR]

Information technology (IT) innovation development within non-IT firms has been a key interest, but it is fraught with challenges because these firms lack sufficient IT knowledge. This study takes a resource dependence perspective to examine how engaging interlocking directorates with IT firms, or IT interlocks, affects non-IT firms' innovation outcomes. Despite the acknowledged role of board interlocks in knowledge transfer, the role of IT interlocks in transferring IT knowledge has not been studied. Using a large-scale panel dataset of Chinese public firms between 2000 and 2020, our findings reveal that IT interlocks of non-IT firms positively impact their IT innovation by transferring IT knowledge, particularly when the interlocked IT firms are knowledge-intensive. Our research contributes to the information systems literature by affirming IT interlocks' positive impact on innovation outcomes and highlighting the value of specific board relational capital in transferring external knowledge in need. It also offers practical implications for non-IT firms overcoming innovation challenges by establishing directorate connections with IT firms. [ABSTRACT FROM AUTHOR]

This paper evaluates the productivity impacts and the subsequent adjustment costs associated with hiring different knowledge workers. I focus on the difference between hiring former entrepreneurs, employees who change jobs, and unemployed individuals. I am the first to evaluate the direct impact that hiring former entrepreneurs has on firm productivity and the heterogenous adjustment costs associated with the different types of new hires. I find no difference between the first-year adjustment costs of entrepreneurs and those of regular-wage employees. Hiring former entrepreneurs is a way to increase productivity after the first year of employment only if the former entrepreneurs are from the highest end of the ability distribution. [ABSTRACT FROM AUTHOR]

Interorganizational partnerships can spur innovation, but their value may be diminished by friction in knowledge flows between firms. We consider how a partner's organizational structure may influence the knowledge that is accessible via partnerships. We focus on how a partner's structure trades off localized autonomy for its managers, which facilitates timelier decision making, and unified control, which facilitates integration. By shaping this balance, centralization of decision rights within the partner organization shapes access to its knowledge. Centralized structures generate wide-ranging internal knowledge pathways that enable access to a broader array of a partner's knowledge. However, the reduced managerial autonomy afforded by centralization makes decision making more cumbersome, which constricts the rate of access to a partner's knowledge. We find evidence of this tradeoff in the context of corporate venture capital relationships between incumbents and startups in the pharmaceutical industry. An increase in the incumbent's diversity of knowledge or in the knowledge required by the startup enhances the value of a greater breadth of access, whereas the degree to which the startup can leverage social ties (affinity) or hierarchical fiat (authority) alleviates the costs of a reduced access rate. Each of these features makes an incumbent organization's centralization more valuable to the startup. By highlighting this tension related to centralization, our findings suggest that new firms striving to maximize their partnership benefits may need to carefully consider their partners' internal structures. [ABSTRACT FROM AUTHOR]

We examine the relation between within-firm office changes and audit quality in the United States. Our primary analysis documents a reduction in audit quality, measured using abnormal discretionary accruals and restatements, when the client is transferred to a smaller within-firm office (downsize effect). We are unable to find evidence that clients experience significant improvement in audit quality among transfers to a larger within-firm office (upsize effect). We then condition our sample on the change in the number of public clients of the receiving office to better understand the source of the underlying association. We find that our downsize effect is driven by offices experiencing a decrease in the number of public clients, suggesting that our main association is not entirely the result of resource constraints for the receiving office. We posit that this finding is consistent with audit quality deterioration among within-firm office changes to smaller offices driven, in part, by the receiving office's inability to adequately overcome the knowledge transfer frictions that accompany a move to a new office. Our findings offer empirical evidence on consequences of within-firm office changes and are particularly relevant to regulators and preparers. [ABSTRACT FROM AUTHOR]

The article is devoted to the study of the objective aspect of transferring or collecting data constituting proprietary information collected in the course of operational and investigative, counterintelligence activities in the field of defense of the country. It is established that the objective side of this criminal offense is expressed in two forms: 1) transfer of data constituting proprietary information collected in the course of operational and investigative, counterintelligence activities in the field of defense of the country to foreign enterprises, institutions, organizations or their representatives; 2) collection of such data for the purpose of transferring to foreign enterprises, institutions, organizations or their representatives. The transfer takes place when a person possesses them and communicates (transmits, delivers) them to foreign enterprises, institutions, organizations or their representatives. The methods of transmission may be different (oral, written, direct familiarization with any materials, transmission via the Internet, by telephone, using caches, couriers, etc.) It also does not matter whether the transferred sources are copies or only information about them. The collection of data constituting official information collected in the course of operational, investigative, counterintelligence activities in the field of national defense should be understood as active actions to search for, concentrate (summarize, generalize) and record (image, photograph, copy, memorize) information. The peculiarity of the transfer and collection of information constituting proprietary information collected in the course of operational and investigative, counterintelligence, and defense activities is that it becomes the property of only foreign enterprises, institutions, organizations, or their representatives, and not of any interested person. Part 1 of Art. 330 of the Criminal Code of Ukraine does not provide for the occurrence of socially dangerous consequences, and, therefore, is not a mandatory feature of the objective side of this criminal offense. The criminal offense under Article 330 of the CC of Ukraine is considered completed either from the moment of transmission or from the moment of collection of information for the purpose of its transmission. The content of the objective side of this criminal offense includes its optional features - method, means, place, time and setting of the commission. These features are a set of specific conditions under which a criminal offense is committed, the objective side develops and a criminal offense result occurs. [ABSTRACT FROM AUTHOR]

As a vital problem in pattern analysis and machine intelligence, Unsupervised Domain Adaptation (UDA) attempts to transfer an effective feature learner from a labeled source domain to an unlabeled target domain. Inspired by the success of the Transformer, several advances in UDA are achieved by adopting pure transformers as network architectures, but such a simple application can only capture patch-level information and lacks interpretability. To address these issues, we propose the Domain-Transformer (DoT) with domain-level attention mechanism to capture the long-range correspondence between the cross-domain samples. On the theoretical side, we provide a mathematical understanding of DoT: (1) We connect the domain-level attention with optimal transport theory, which provides interpretability from Wasserstein geometry; (2) From the perspective of learning theory, Wasserstein distance-based generalization bounds are derived, which explains the effectiveness of DoT for knowledge transfer. On the methodological side, DoT integrates the domain-level attention and manifold structure regularization, which characterize the sample-level information and locality consistency for cross-domain cluster structures. Besides, the domain-level attention mechanism can be used as a plug-and-play module, so DoT can be implemented under different neural network architectures. Instead of explicitly modeling the distribution discrepancy at domain-level or class-level, DoT learns transferable features under the guidance of long-range correspondence, so it is free of pseudo-labels and explicit domain discrepancy optimization. Extensive experiment results on several benchmark datasets validate the effectiveness of DoT. [ABSTRACT FROM AUTHOR]

Multi-label image recognition with partial labels (MLR-PL), in which some labels are known while others are unknown for each image, may greatly reduce the cost of annotation and thus facilitate large-scale MLR. We find that strong semantic correlations exist within each image and across different images, and these correlations can help transfer the knowledge possessed by the known labels to retrieve the unknown labels and thus improve the performance of the MLR-PL task. In this work, we propose a novel heterogeneous semantic transfer (HST) framework that consists of two complementary transfer modules that explore both within-image and cross-image semantic correlations to transfer the knowledge possessed by known labels to generate pseudo labels for the unknown labels. Specifically, an intra-image semantic transfer (IST) module learns an image-specific label co-occurrence matrix for each image and maps the known labels to complement the unknown labels based on these matrices. Additionally, a cross-image transfer (CST) module learns category-specific feature-prototype similarities and then helps complement the unknown labels that have high degrees of similarity with the corresponding prototypes. It is worthy-noting that the HST framework requires searching appropriate thresholds to determine the co-occurrence and similarity scores to generate pseudo labels for the IST and CST modules, respectively. To avoid highly time-consuming and resource-intensive manual tuning, we introduce a differential threshold learning algorithm that adjusts the nondifferential indication function to a differential formulation to automatically learn the appropriate thresholds. Finally, both the known and generated pseudo labels are used to train MLR models. Extensive experiments conducted on the Microsoft COCO, Visual Genome, and Pascal VOC 2007 datasets show that the proposed HST framework achieves superior performance to that of current state-of-the-art algorithms. Specifically, it obtains mean average precision (mAP) improvements of 1.4, 3.3, and 0.4% on the three datasets over the results of the best-performing previously developed algorithm. [ABSTRACT FROM AUTHOR]

Cross-Domain Few-Shot Learning (CD-FSL) aims at recognizing unseen classes from target domains that vastly differ from training classes from source domains, utilizing only a few labeled samples. However, the substantial domain disparities between target and source domains pose huge challenges to few-shot generalization. To resolve domain disparities, we propose HybridPrompt, a novel architecture for Domain-Aware Prompting that integrates a variety of cross-domain learned prompts as knowledge experts for CD-FSL. The proposed method enjoys several merits. First, to encode knowledge from diverse source domains, several Domain Prompts are introduced to capture domain-specific knowledge. Subsequently, to facilitate the cross-domain transfer of valuable knowledge, a Transferred Prompt is specifically tailored for each target task by retrieving highly relevant Domain Prompts based on domain properties. Finally, to complement insufficient transferred information, an Adaptive Prompt is learned to incorporate additional target characteristics for model adaptation. Consequently, the collaboration of these three types of prompts contributes to a hybridly prompted model that achieves domain-aware encoding, transfer, and adaptation, thereby enhancing adaptability on unseen domains. Extensive experimental results on the Meta-Dataset benchmark demonstrate that our method achieves superior performance against state-of-the-art methods. The source code is available at https://github.com/Jamine-W/HybridPrompt. [ABSTRACT FROM AUTHOR]

Employing the Foucauldian term 'conduct', this article explores how social resilience and morale became a target of state intervention in Denmark during the Cold War. 'Psychological defence' was a Cold War phenomenon designed to bring an imagined future war into a space of control as well as a tool for the authorities' exercise of power in case another world war became a reality. Advocating a methodological internationalism, the article analyses how the concept of psychological defence travelled from Sweden to Denmark via Norway and NATO, and in a complex process of translation, mixing and hybridization was adapted and appropriated to Danish security policy conditions, preparedness culture, and historical experiences. Ultimately, psychological defence was replaced with a more practical or even cynical approach to public information and media preparedness, even if the objectives remained the same. The article employs source material from Danish, Swedish, and NATO archives and combines Scandinavian Cold War history with media history and the history of knowledge. [ABSTRACT FROM AUTHOR]

The weight initialization technique for transfer learning refers to the practice of using pretrained models that can be modified to solve new problems, instead of starting the training process from scratch. In this study, six different transfer learning weight initialization strategies were proposed for plant disease detection: scratch (i.e., random initialization), pretrained model on cross‐domain (ImageNet), model trained on related domain (ISIC 2019), model trained on related domain (ISIC 2019) with cross‐domain (ImageNet) weights, model trained on same domain (PlantVillage), and model trained on same domain (PlantVillage) with cross‐domain weights (ImageNet). Weights from each strategy were transferred to a target dataset (Plant Pathology 2021). These strategies were implemented using eight deep learning architectures. It was observed that transferring from any strategy led to an average acceleration of convergence ranging from 33.88% to 73.16% in mean loss and an improvement of 8.72%–42.12% in mean F1‐score compared to the scratch strategy. Moreover, although smaller and less comprehensive than ImageNet, transferring information from the same domain or related domain proved to be competitive compared to transferring from ImageNet. This indicates that ImageNet, which is widely favoured in the literature, may not necessarily represent the most optimal transfer source for the given context. In addition, to identify which strategies have significant differences, a post hoc analysis using Tukey's HSD test was conducted. Finally, the classifications made by the proposed models were visualized using Grad‐CAM to provide a qualitative understanding of how different weight initialization strategies affect the focus areas of the models. [ABSTRACT FROM AUTHOR]

Unsupervised domain adaptation (UDA) aims to reduce the distribution discrepancy across domains, enabling the transfer of knowledge from the labeled source domain to the unlabeled target domain. The main focus of most current UDA methods lies on extracting domain invariant representations to reduce the gap between domains. However, this singular emphasis on domain invariance may inadvertently ignore domain relevant information, which may lead to negative transfer effects. Moreover, current adversarial DA methods give uniform weight to all samples, overlooking the distinct influence that variations within source domain data or noise may exert on the adversarial performance. To address these challenges, we propose a novel method called Mutual-weighted Feature Disentanglement for Unsupervised Domain Adaptation (MWFD). Specifically, we decouple domain invariant features from domain specific features, and then use the entropy of the classifier to rebalance the weights of the domain discriminator, and simultaneously adjust the weights of the classifier using the domain entropy of the domain discriminator to reduce domain discrepancies. Finally, to obtain more discriminative features, we adopt a self-supervised contrastive learning framework to bring positive sample pairs closer together while pushing negative sample pairs apart, enhancing the discriminability of the model on the target domain. Extensive experiments on five benchmark datasets demonstrate that our model outperforms several state-of-the-art domain adaptation methods. [ABSTRACT FROM AUTHOR]

Cloud application programming interface (API) is a software intermediary that enables applications to communicate and transfer information to one another in the cloud. As the number of cloud APIs continues to increase, developers are inundated with a plethora of cloud API choices, so researchers have proposed many cloud API recommendation methods. Existing cloud API recommendation methods can be divided into two types: content-based (CB) cloud API recommendation and collaborative filtering-based (CF) cloud API recommendation. CF methods mainly consider the historical information of cloud APIs invoked by mashups. Generally, CF methods have better recommendation performances on head cloud APIs due to more interaction records, and poor recommendation performances on tail cloud APIs. Meanwhile, CB methods can improve the recommendation performances of tail cloud APIs by leveraging the content information of cloud APIs and mashups, but their overall performances are not as good as those of CF methods. Moreover, traditional cloud API recommendation methods ignore the complementarity relationship between mashups and cloud APIs. To address the above issues, this paper first proposes the complementary function vector (CV) based on tag co-occurrence and graph convolutional networks, in order to characterize the complementarity relationship between cloud APIs and mashups. Then we utilize the attention mechanism to systematically integrate CF, CB, and CV methods, and propose a model named Content and Complementarity enhanced Attentional Collaborative Filtering (CCeACF). Finally, the experimental results show that the proposed approach outperforms the state-of-the-art cloud API recommendation methods, can effectively alleviate the long tail problem in the cloud API ecosystem, and is interpretable. [ABSTRACT FROM AUTHOR]

Although innovation is present in all economic sectors, innovation surveys and empirical research are biased towards high-tech activities. Meanwhile, quantitative studies on agricultural innovation systems (AIS) usually neglect the regional dimension. As in many developing countries, Argentine agriculture seems to have evolved towards a new techno-productive paradigm, but no study has yet analyzed its geographical scope. This article proposes a regional-structural approach to study where, how and with whom technical linkages and interactive learning processes are developed. Using under-explored data from the 2018 National Agricultural Census and multivariate analysis techniques, the results reveal significant regional differences in technical linkages and show that the new techno-productive paradigm is limited to the Pampean region. While private sources of knowledge are mainly concentrated in central areas, technical linkages with public agencies are geographically widespread. Acknowledging these regional heterogeneities poses different policy challenges for promoting AIS in developing countries with large and diverse territories. [ABSTRACT FROM AUTHOR]

Regional enlargement and Transit-Oriented Development (TOD) has in Sweden been an active policy and strategy at both municipal, regional, and national level to make society more resource efficient and strengthen local labour markets. The paper aims to explore and discuss how strategic local-regional planning can be used to counteract local, municipal and sub-regional uneven development enhanced by city-centric transit-oriented development. Mutual learning processes between local and regional actors can be used to benefit regional investments and manage local expectations. This means including local knowledge of the local situation and local knowledge of the regional context in the regional planning process and to institutionalize local-regional integration and common projections of TOD. [ABSTRACT FROM AUTHOR]

This study examines whether media co-coverage – a phenomenon where multiple firms are simultaneously mentioned in the same news article as contextual information – induces excessive inter-industry information transfers between two firms due to the increased saliency of their relationship. Using firms from different product market industries that are co-covered in the same Wall Street Journal article, we find that, after co-coverage, the stock price of a co-covered focal firm reacts positively to the earnings surprise of another early-announcing co-covered peer, followed by a reversal on the focal firm's subsequent earnings announcement day, while there is no reaction to the peer's earnings news in the pre-co-coverage period. Further analysis suggests that the transfer and the reversal are stronger when the co-coverage information is more salient to investors, and are concentrated among firms with more active retail trading. These findings suggest that co-coverage in financial media, through the saliency effect, can lead to inefficient cross-industry information transfers. [ABSTRACT FROM AUTHOR]

With the rapid growth of big data, extracting meaningful knowledge from data is crucial for machine learning. The existing Swarm Learning data collaboration models face challenges such as data security, model security, high communication overhead, and model performance optimization. To address this, we propose the Swarm Mutual Learning (SML). Firstly, we introduce an Adaptive Mutual Distillation Algorithm that dynamically controls the learning intensity based on distillation weights and strength, enhancing the efficiency of knowledge extraction and transfer during mutual distillation. Secondly, we design a Global Parameter Aggregation Algorithm based on homomorphic encryption, coupled with a Dynamic Gradient Decomposition Algorithm using singular value decomposition. This allows the model to aggregate parameters in ciphertext, significantly reducing communication overhead during uploads and downloads. Finally, we validate the proposed methods on real datasets, demonstrating their effectiveness and efficiency in model updates. On the MNIST dataset and CIFAR-10 dataset, the local model accuracies reached 95.02% and 55.26%, respectively, surpassing those of the comparative models. Furthermore, while ensuring the security of the aggregation process, we significantly reduced the communication overhead for uploading and downloading. [ABSTRACT FROM AUTHOR]

Robots can use auditory, visual, or haptic interfaces to convey information to human users. The way these interfaces select signals is typically pre-defined by the designer: for instance, a haptic wristband might vibrate when the robot is moving and squeeze when the robot stops. But different people interpret the same signals in different ways, so that what makes sense to one person might be confusing or unintuitive to another. In this article, we introduce a unified algorithmic formalism for learning co-adaptive interfaces from scratch. Our method does not need to know the human's task (i.e., what the human is using these signals for). Instead, our insight is that interpretable interfaces should select signals that maximize correlation between the human's actions and the information the interface is trying to convey. Applying this insight we develop Learning Interfaces to Maximize Information Transfer (LIMIT). LIMIT optimizes a tractable, real-time proxy of information gain in continuous spaces. The first time a person works with our system the signals may appear random; but over repeated interactions, the interface learns a one-to-one mapping between displayed signals and human responses. Our resulting approach is both personalized to the current user and not tied to any specific interface modality. We compare LIMIT to state-of-the-art baselines across controlled simulations, an online survey, and an in-person user study with auditory, visual, and haptic interfaces. Overall, our results suggest that LIMIT learns interfaces that enable users to complete the task more quickly and efficiently, and users subjectively prefer LIMIT to the alternatives. See videos here: https://youtu.be/IvQ3TM1_2fA. [ABSTRACT FROM AUTHOR]

Firms can derive insights not only from their own failures but also from the innovation failures of counterparts in the same industry. This study investigates how 109 Chinese A‐share listed high‐tech firms (2013–2022) learn from counterparts' innovation failures to improve their success rate of innovation and emphasizes the crucial role of venture capital (VC) syndication networks in facilitating knowledge transfer. Central VCs within the network notably enhance vicarious learning from others' failures. Further empirical analysis shows a significant increase in the success rate of exploitative innovations, while the impact on exploratory innovations is minimal. [ABSTRACT FROM AUTHOR]

Dehazing can improve the clarity of images and provide more reliable inputs for image analysis tasks, thereby enhancing their performance. Therefore, we propose a dehazing network based on knowledge transfer and multi-data enhancement correction. First, we propose a multi-data enhancement correction method that combines different image enhancement techniques to improve the quality of the input images. Second, by leveraging a pre-trained teacher network to acquire prior knowledge from clear data, guiding the dehazing process of the student network through knowledge transfer. We introduce a deep multi-scale refinement network composed of a dense feature enhancement module and enhanced residual dense blocks, enabling the dehazing model to learn the local structure and feature representation of the data more accurately. Experimental results on multiple benchmark datasets demonstrate that the proposed dehazing method outperforms state-of-the-art dehazing methods. Code are available at: https://github.com/JNcmm/KTMDA-DehazeNet. [ABSTRACT FROM AUTHOR]

The article offers an examination of augmented reality (AR)-assisted disaster preparedness guidelines designed to improve public awareness and engagement. Topics include the use of AR to enhance disaster preparedness by incorporating object recognition for environmental hazard identification, a comparison of AR-based guidelines with traditional paper-based ones in terms of usability and task performance, and the findings from testing these interfaces across different age groups.

Group-living organisms commonly exhibit collective escape responses, yet how information flows among group members in these events remains an open question. Here, we study the collective responses of a sheep flock (Ovis aries) to a shepherd dog (border collie) in a driving task between two well-defined target points. We collected high-resolution spatiotemporal data from 14 sheep and the dog, using Ultra-Wide-Band tags attached to each individual. We find that the spatial positions of sheep along the front-back axis of the group's velocity strongly correlate with their impact on the collective movement. Our analyses reveal that, even though the dog chases the sheep flock from behind, directional information on shorter time scales propagates from the front of the group towards the rear; further, the dog adjusts its movement in response to the flock's dynamics. We introduce an agent-based model that captures key data features. Specifically, in response to chasing, the sheep change their spatial relative positions less frequently and exhibit a transfer of directional information flow from front to back; this pattern disappears in the absence of chasing. Our study reveals some general insights into how directional information propagates in escaping animal groups. Experimental analysis and computational modeling of collective responses of flocking sheep to a herding dog reveal how directional information spreads in escaping animal groups. [ABSTRACT FROM AUTHOR]

Classification of epileptic electroencephalograms can provide strong technical support for early warning and monitoring of epilepsy. However, Electroencephalogram (EEG) has characteristics of specificity, multi-feature, and susceptibility to noise interference. Therefore, this paper proposes a joint low-rank representation and transfer multi-view learning (LRR-TML) approach for EEG epilepsy classification. The LRR-TML approach inherits the advantages of low-rank representation, filters redundant features and noise information, and focuses on the main feature and significant feature of EEG. The LRR-TML approach introduces the ideas of transfer learning and multi-view learning on the framework of low-rank representation, mines the linear correlation between different domains in the subspace through low-rank representation, and effectively transfers the discriminative knowledge of the source domain to the target domain. According to the principles of complementarity and consistency of views, the LRR-TML approach comprehensively describes EEG by integrating various features, maximizing the consistency between different views and minimizing the divergence between views. Experimental results on the Bonn dataset show that the LRR-TML approach achieves the best classification results among all comparison approaches. [ABSTRACT FROM AUTHOR]

The outbreak of the COVID-19 pandemic caused dashboards to become widely used by the public and decision-makers. Nevertheless, dashboard interfaces have been related to business intelligence since their origins, and the search for improvements in their design is not new. This article's objective is to conduct a user evaluation of COVID-19 dashboards that contain geospatial information. This is done through a formative study to identify problematic aspects of user/dashboard interaction. This is enhanced by comparing two self-developed dashboards that, according to previous tests, have functionalities with different appearances. User evaluation is performed through mixed research that combines objective (eye-tracking) and subjective (a questionnaire and an interview) methods. The results generate recommendations for better-designed dashboard interfaces that can transfer information appropriately. The vital elements needed to achieve this are interactivity, the option to choose the metrics, and the distribution of the elements in the layout, all playing a role in a more user-friendly interaction between the user and the dashboard. [ABSTRACT FROM AUTHOR]

A striking feature of human cognition is an exceptional ability to rapidly adapt to novel situations. It is proposed this relies on abstracting and generalizing past experiences. While previous research has explored how humans detect and generalize single sequential processes, we have a limited understanding of how humans adapt to more naturalistic scenarios, for example, complex, multisubprocess environments. Here, we propose a candidate computational mechanism that posits compositional generalization of knowledge about subprocess dynamics. In two samples (N = 238 and N = 137), we combined a novel sequence learning task and computational modeling to ask whether humans extract and generalize subprocesses compositionally to solve new problems. In prior learning, participants experienced sequences of compound images formed from two graphs' product spaces (group 1: G1 and G2, group 2: G3 and G4). In transfer learning, both groups encountered compound images from the product of G1 and G3, composed entirely of new images. We show that subprocess knowledge transferred between task phases, such that in a new task environment each group had enhanced accuracy in predicting subprocess dynamics they had experienced during prior learning. Computational models utilizing predictive representations, based solely on the temporal contiguity of experienced task states, without an ability to transfer knowledge, failed to explain these data. Instead, behavior was consistent with a predictive representation model that maps task states between prior and transfer learning. These results help advance a mechanistic understanding of how humans discover and abstract subprocesses composing their experiences and compositionally reuse prior knowledge as a scaffolding for new experiences. [ABSTRACT FROM AUTHOR]

Based on the study of the professional careers of 439 parliamentary assistants of the 7th parliamentary term (2009–2014), we identify their different career patterns, especially after leaving the European parliament. We show which factors (social, biographical) influence mobility within the field of Eurocracy. We assess the importance of public/private circulation. Beyond the notorious ‘revolving doors’ and the question of the transfer of information from EU institutions to private interests, we show that parliamentary assistants form a group of specialised intermediaries of EU affairs who, after having acquired specific knowledge and know-how at the heart of European policymaking, are able to convert these assets in different sectors (public bureaucracies, private companies, or NGOs), at the EU level or in the member states. Such mobility illustrates the blurring of the public/private divide, and the hybridisation of EU governance. We thus shed new light on the structuring of the field of Eurocracy and the circulation of professionals within it. [ABSTRACT FROM AUTHOR]

Timely detection of hard exudates in fundus images can effectively avoid the severity of the disease, but the labelling of small and numerous lesion areas requires a lot of labour costs. This paper proposes a semi‐supervised segmentation network, which integrates pseudo‐labels and multi‐level features consistency correction. It achieves accurate segmentation of hard exudates by making full use of a small amount of labelled data and a large amount of unlabelled data. The network effectively extracts features from the unlabelled data through knowledge transfer of the teacher‐student model, and incorporates a Transformer network for auxiliary training to promote the quality of transfer. In addition, three unsupervised losses are introduced to improve the performance: the perturbation loss improves the robustness of the model to noise by adding different noises to the same input; the multi‐level feature consistency correction loss ensures the consistency of features of the student model at different scales; and the pseudo‐labelling cross‐supervision loss utilizes the generated pseudo‐labels for supervision between CNN and Transformer. By comparing the segmentation results with different proportion of the labelled data, it has better segmentation performance compared to other methods. The proposed methods can totally increase dice by 16.56% and mean intersection over union (MIoU) by 25.11%. [ABSTRACT FROM AUTHOR]

Social network–based data collection methods that rely on third‐party reporting have emerged as a promising approach for measuring abortion in restrictive settings. In order for these methods to accurately measure abortion incidence, they must also assess the visibility of abortions within social networks. Failure to do so may result in estimates affected by transmission bias, caused by imperfect knowledge of all abortions within one's social network. In this paper, we present exploratory research that uses respondent‐driven sampling (RDS) and the game of contacts method to measure abortion visibility in four sites in Uganda and Ethiopia. We assess the existence of potential biases in the game of contacts estimate of abortion visibility in each site by conducting several internal and external validity tests. While these tests provided some promising results, other factors such as the representativeness of the RDS samples, direct versus indirect abortion knowledge transfers, and the generalizability of the study sites may have introduced biases into the final estimates of abortion visibility in this study. We conclude by making recommendations on how applications of this methodology could be improved to better estimate abortion‐related transmission bias. [ABSTRACT FROM AUTHOR]

The creation of successful innovation ecosystems, like Silicon Valley, is challenging due to significant cultural, infrastructural, and resource differences between regions. In this context, the Innovation Reef (I-Reef) model emerges as a promising alternative, offering an approach for regions with limited resources to develop successful innovation ecosystems based on cooperation and mutual benefit among participants. This model has great potential to promote regional development, especially due to its focus on retaining and sharing the value generated. However, the role of I-Reef in sustainable regional development still needs to be further explored. Thus, the objective of this study is to deepen the theoretical understanding of the I-Reef model by analyzing its contribution to sustainable development. To achieve this, a comparison was made between I-Reef and established models such as business, innovation, knowledge, and entrepreneurial ecosystems. A systematic literature review conducted on Scopus found 704 articles published in the last three decades. The purpose was to identify the similarities and differences between the models of innovation business ecosystem models. The results show that there is alignment between I-Reef and the different ecosystems on several points. A central aspect of I-Reef is that it relies on a strong network of mutually beneficial relationships, much more oriented to sustainable development than the other models, which is a key factor in generating competitive advantage and development for the region. This characteristic is either not addressed or not placed at the core of the ecosystems discussed in the literature. For future research, empirical studies and validation of the I-Reef model with experts are suggested, as this theoretical study lays the foundation for more in-depth analyses. [ABSTRACT FROM AUTHOR]

Surrogate-assisted evolutionary algorithms have been widely employed to solve data-driven optimization problems. However, for offline data-driven optimization, it is very challenging to perform evolutionary search efficiently as well as accurately since no new data is available during the optimization process. To mitigate this issue, a multifidelity surrogates-assisted multi-objective particle swarm optimization (MFSa-PSO) algorithm is proposed in this paper. First, two low-fidelity models with convergence and diversity characteristics separately and a high-fidelity model are constructed to assemble multifidelity surrogate models. Second, by adopting the knowledge transfer strategy, the multifidelity surrogates-assisted two-archive multi-objective particle swarm optimization is conducted to search optimal solutions more exactly and effectively. Third, the output solution set is achieved by associating the solutions of two archives with reference vectors. Finally, the proposed MFSa-PSO is compared with some popular surrogate-assisted evolutionary algorithms on benchmark problems to verify its effectiveness and outperformance. Additionally, a real-world application of the municipal solid waste incineration process is carried out to verify the engineering applicability of MFSa-PSO. [ABSTRACT FROM AUTHOR]

Background: Electroencephalogram (EEG) is widely used in emotion recognition due to its precision and reliability. However, the nonstationarity of EEG signals causes significant differences between individuals or sessions, making it challenging to construct a robust model. Recently, domain adaptation (DA) methods have shown excellent results in cross-subject EEG emotion recognition by aligning marginal distributions. Nevertheless, these methods do not consider emotion category labels, which can lead to label confusion during alignment. Our study aims to alleviate this problem by promoting conditional distribution alignment during domain adaptation to improve cross-subject and cross-session emotion recognition performance. Method: This study introduces a multi-source domain adaptation common-branch network for EEG emotion recognition and proposes a novel sample hybridization method. This method enables the introduction of target domain data information by directionally hybridizing source and target domain samples without increasing the overall sample size, thereby enhancing the effectiveness of conditional distribution alignment in domain adaptation. Cross-subject and cross-session experiments were conducted on two publicly available datasets, SEED and SEED-IV, to validate the proposed model. Result: In cross-subject emotion recognition, our method achieved an average accuracy of 90.27% on the SEED dataset, with eight out of 15 subjects attaining a recognition accuracy higher than 90%. For the SEED-IV dataset, the recognition accuracy also reached 73.21%. Additionally, in the cross-session experiment, we sequentially used two out of the three session data as source domains and the remaining session as the target domain for emotion recognition. The proposed model yielded average accuracies of 94.16 and 75.05% on the two datasets, respectively. Conclusion: Our proposed method aims to alleviate the difficulties of emotion recognition from the limited generalization ability of EEG features across subjects and sessions. Though adapting the multi-source domain adaptation and the sample hybridization method, the proposed method can effectively transfer the emotion-related knowledge of known subjects and achieve accurate emotion recognition on unlabeled subjects. [ABSTRACT FROM AUTHOR]

Data visualization is a complex matter and so is teaching it in the context of engineering and data science. Even more challenging is teaching in a midst of once in a century pandemic. In this paper, we show the results of a workshop conducted with second-year students on a university degree in data science, which took place in the context of pandemic restrictions in a hybrid learning environment. About half of the students were in the classroom and the other half at home. They learned to use STMaps, a tool developed in the scope of the SILKNOW project. We transferred research outcomes to the classroom; similarly, the results of evaluating the tool were transferred to the project. But more importantly, we found insights that the learning process of students that were online was not the same quality as of those that were physically in the classroom. While both groups performed well in terms of correctly fulfilling the tasks proposed during the workshop, the students at home were less motivated and perceived the tasks more cumbersome to achieve. From a research perspective, this activity was perceived as very useful, and similar activities are being planned for each academic year. [ABSTRACT FROM AUTHOR]

Background Methods Results Conclusion The American Speech–Language–Hearing Association (ASHA) promotes the use of evidence‐based practice (EBP) in the provision of clinical services by speech–language pathologists (SLPs). Therefore, training in the EBP process is vital in the education of future SLPs. Case‐based learning (CBL) has been found effective in facilitating transfer of knowledge and skills in various fields; however, its utilisation to teach EBP to SLP students is understudied.This phenomenological investigation explores the use of CBL, an active learning strategy that gives students the opportunity to practice application of knowledge to realistic scenarios, to teach EBP to graduate SLP students. Graduate SLP students worked in small groups to complete an article review activity centred around an assigned hypothetical clinical case. Students submitted article review papers and written reflections on their learning process. The written reflections and review papers were analysed using qualitative content analysis to understand students' approach to clinical cases and their perspectives on the CBL approach.Analysis of the papers and reflections indicated that students tended to rely heavily on scientific evidence for clinical decision making, often to the exclusion of clinical expertise or client perspectives. Students reported that the project was applicable to their future clinical practice. Further, they reported learning from multiple sources, including the instructor, class discussion and their peers.Results indicate that graduate SLP students found CBL activities beneficial for learning the principles of EBP. Students specifically indicated an appreciation for the realistic nature of the activities. Future applications are examined. [ABSTRACT FROM AUTHOR]

In 1492, a Lübeckian printer Stephan Arndes released an updated edition of a popular late medieval hagiographic collection Der Heiligen Leben in Middle Low German. From a few dozens of newly added legends, nine of them are about saints who were venerated only in Northern Germany and Scandinavia – or both. By adding these ‘Northern’ hagiographies, Arndes attempted to domesticate an otherwise rather universal collection – and not only among a Northern German but also a Scandinavian audience. By analysing these hagiographies, their possible Latin sources, and how Stephan Arndes got hold of these texts, the article shows that the edition in question is a testimony to transregional hagiographic knowledge transfer between Northern Germany and Scandinavia. Such an exchange was realized through specific editorial decisions, printing networks, and business connections, enabled by the wide usage of Middle Low German in the Nordic countries not only among foreign merchants but also locals, especially in legal and religious milieux. This case study of Stephan Arndes’ hagiographic print reveals that late medieval transregional devotional networks were bi-directional, unlike what is suggested by the centre-periphery paradigm, often applied when discussing medieval German influences on Scandinavian religious culture and institutions. [ABSTRACT FROM AUTHOR]

The tree of blobs of a species network shows only the tree-like aspects of relationships of taxa on a network, omitting information on network substructures where hybridization or other types of lateral transfer of genetic information occur. By isolating such regions of a network, inference of the tree of blobs can serve as a starting point for a more detailed investigation, or indicate the limit of what may be inferrable without additional assumptions. Building on our theoretical work on the identifiability of the tree of blobs from gene quartet distributions under the Network Multispecies Coalescent model, we develop an algorithm, TINNiK, for statistically consistent tree of blobs inference. We provide examples of its application to both simulated and empirical datasets, utilizing an implementation in the MSCquartets 2.0 R package. [ABSTRACT FROM AUTHOR]

Sensory data are the basis for the intelligent health state awareness of planetary gearboxes, which are the critical components of electromechanical systems. Despite the advantages of intelligent diagnostic techniques for detecting intricate fault patterns and improving diagnostic speed, challenges still persist, which include the limited availability of fault data, the lack of labeling information and the discrepancies in features across different signals. Targeting this issue, a subdomain distribution adversarial adaptation diagnosis method (SDAA) is proposed for faults diagnosis of planetary gearboxes across different conditions. Firstly, nonstationary vibration signals are converted into a two-dimensional time–frequency representation to extract intrinsic information and avoid frequency overlapping. Secondly, an adversarial training mechanism is designed to evaluate subclass feature distribution differences between the source and target domain. A conditional distribution adaptation is employed to account for correlations among data from different subclasses. Finally, the proposed method is validated through experiments on planetary gearboxes, and the results demonstrate that SDAA can effectively diagnose faults under crossing conditions with an accuracy of 96.7% in diagnosing gear faults and 95.2% in diagnosing planet bearing faults. It outperforms other methods in both accuracy and model robustness. This confirms that this approach can refine domain-invariant information for transfer learning with less information loss from the sub-class level of fault data instead of the overall class level. [ABSTRACT FROM AUTHOR]

This study used a quasirandomized within-classroom design to investigate whether prior knowledge about physics gained in elementary school prepares students for future learning in related content areas in secondary school. A total of 433 children (intervention group) received four basic curriculum units on physics from their elementary school teachers. The units dealt with floating and sinking, air and atmospheric pressure, the stability of bridges, and sound and the spreading of sound. These children entered 60 newly composed classes in early secondary school that completed an advanced curriculum unit on hydrostatic pressure and buoyancy force with their secondary school teachers. A total of 942 students (control group) in these classes had not received the four basic physics curriculum units. On a conceptual knowledge test about hydrostatic pressure and buoyancy force, the intervention group outperformed the control group in the pretest (d = 0.28) and in the posttest (d = 0.25). Students in the intervention group showed similar learning gains as those in the control group, but when controlling for pretest performance, they achieved higher learning outcomes. Regression analyses within the intervention group revealed that this advantage resulted from the content-specific transfer of conceptual knowledge from topically related basic curriculum units. The basic physics instruction also prepared male and female students equally for future learning. Educational Impact and Implications Statement: Our study shows that students who have received early physics instruction in elementary school benefit slightly more from later, more advanced physics instruction in secondary school. This finding provides first evidence that the idea of a spiral curriculum, in which learners first build basic knowledge that is later on expanded in more demanding instruction, can work, although it is yet to be further examined how this process can be optimized in school instruction. [ABSTRACT FROM AUTHOR]

Social foraging is commonly used across taxa to increase animal foraging success in uncertain environments and is believed to be a driver of social group formation. In temperate zones, females of many bat species form seasonal colonies, whereas males are usually solitary. Males of only a few bat species form temporary colonies during sperm production, likely to benefit from social foraging, social thermoregulation, or both. We radiotracked a group of reproductive male particoloured bats, Vespertilio murinus , to test the hypothesis that they use social foraging. Foraging bats overlapped in time and space significantly more than expected by chance, suggesting that they used social information to increase detection of insect swarms. Dyads also sometimes switched foraging patches together, suggesting part-time use of the more coordinated group facilitation social foraging strategy. Our results support the hypothesis that male particoloured bats use local enhancement mixed with group facilitation during sperm production and that improved foraging success through information transfer is a likely driver of seasonal sociality in these and other male bats. • Male particoloured bats foraged socially. • Bats often changed their direction of flight at the same time. • Males did not forage together during all nocturnal activity. • Bats probably used local enhancement mixed with group facilitation. [ABSTRACT FROM AUTHOR]

Immersive virtual reality (VR) is anticipated to peak in development this decade bringing new opportunities for 3D multimodal designing across all levels of education. The need for students to gain capabilities with multimodal texts—texts that combine two or more modes, such as spoken, written, and visual—is emphasised at all levels of education from P-12 in the Australian Curriculum. Likewise, the use of technology-supported pedagogies is increasing worldwide, rendering multimodal texts ubiquitous across all knowledge domains. This original, qualitative classroom research investigated students' 3D designing of multimodal texts using an immersive VR head-mounted display. Upper primary students (ages 10–12 years, n = 48) transferred their knowledge of ancient Rome through 2D drawing, writing, speaking, and 3D multimodal designing with VR. The application of multimodal analysis to video data, screen recordings, and think-aloud protocols, and the thematic coding of student and teacher interviews yielded four key findings: (i) VR gaming supported 3D multimodal designing through haptic and embodied experience, (ii) VR improved performance through creative redesigning, (iii) VR-supported knowledge application, consolidation, and transfer, and (iv) pedagogical strengths of VR were situated and transformed practice. This research is timely and significant given the increasing accessibility and affordability of VR and the need to connect research and pedagogical practice to support students' advanced knowledge and capabilities with multimodal learning across the curriculum. [ABSTRACT FROM AUTHOR]

Approximately 70%–75% of people worldwide have no formally registered land rights. Fit-For-Purpose Land Administration was introduced to address this problem and focuses on delineating visible cadastral boundaries from earth observation imagery. Recent studies have shown the potential of deep learning models to extract these visible cadastral boundaries automatically. However, studies are limited by the small size and geographical coverage of available datasets and by the lack of information about which cadastral boundaries are visible, i.e., associated with a physical object boundary. To overcome these problems, we present CadastreVision , a benchmark dataset containing cadastral reference data and corresponding multi-resolution earth observation imagery from The Netherlands, with a spatial resolution ranging from 0.1 m to 10 m. The ratio between visible and non-visible cadastral boundaries is essential to evaluate the potential automation level in cadastral boundary extraction from earth observation images and interpret results obtained by deep learning models. We investigate this ratio using a novel analysis pipeline that overlays cadastral reference data with visible topographic object boundaries. Our results show that approximately 72% of the total length of cadastral boundaries in The Netherlands are visible. CadastreVision will enable new developments in cadastral boundary delineation and future endeavours to investigate knowledge transfer to data-scarce areas. Our data and code is available at https://github.com/jeroengrift/cadastrevision. [Display omitted] [ABSTRACT FROM AUTHOR]