Hasil untuk "Academies and learned societies"

Marta Nunes da Costa

O movimento dos ideólogos tem sido negligenciado dentro da da história da filosofia. Este artigo visa colmatar, mesmo que parcialmente, esta lacuna. Para isso, no primeiro momento, oferecemos uma contextualização do movimento dos ideólogos e de sua relação com o enciclopedismo e os philosophes. No segundo momento, voltamo-nos para a primeira geração de ideólogos, estabelecendo o pano de fundo das discussões da época revolucionária. No terceiro momento, apontamos algumas das ideias centrais presentes na obra Elementos de Ideologia de Destutt de Tracy. Terminamos com uma reflexão acerca da importância de resgatar os princípios e o método de bem pensar presentes na obra de Tracy para os dias de hoje. 

Tom Lang

Inclusive language is ‘language free of stereotypes, implicit bias, and negative messages’. The inclusive language movement intends to ‘acknowledge diversity, convey respect to all people, be sensitive to differences, and promote equal opportunities’. However, inclusive language is an idea or a value, not a widespread, organised effort to establish a definitive set of terms. Who decides what terms to use? What are the costs and consequences of establishing these terms? To better understand the movement, I looked at it from the perspective of diffusion theory, which seeks to explain how new products, services, and ideas are adopted (diffused) in a social system over time. The theory has identified five characteristics of successful innovations: 1) high relative advantage over alternatives, 2) high compatibility with personal and social norms, 3) low complexity in adoption and use, 4) high ‘triability’ or the chance to use the innovation before adoption, and 5) high visibility that confirms the choice of adoption. By these characteristics, many inclusive language terms face substantial barriers to widespread voluntary acceptance. These same five characteristics, however, can help inform the movement by identifying which terms are more likely to be accepted. Here, I identify where non-inclusive terms appear in the language and suggest how diffusion theory can be used to assess the likelihood of their adoption.

Kangan Qian, Jinyu Miao, Ziang Luo et al.

Accurate and reliable spatial and motion information plays a pivotal role in autonomous driving systems. However, object-level perception models struggle with handling open scenario categories and lack precise intrinsic geometry. On the other hand, occupancy-based class-agnostic methods excel in representing scenes but fail to ensure physics consistency and ignore the importance of interactions between traffic participants, hindering the model's ability to learn accurate and reliable motion. In this paper, we introduce a novel occupancy-instance modeling framework for class-agnostic motion prediction tasks, named LEGO-Motion, which incorporates instance features into Bird's Eye View (BEV) space. Our model comprises (1) a BEV encoder, (2) an Interaction-Augmented Instance Encoder, and (3) an Instance-Enhanced BEV Encoder, improving both interaction relationships and physics consistency within the model, thereby ensuring a more accurate and robust understanding of the environment. Extensive experiments on the nuScenes dataset demonstrate that our method achieves state-of-the-art performance, outperforming existing approaches. Furthermore, the effectiveness of our framework is validated on the advanced FMCW LiDAR benchmark, showcasing its practical applicability and generalization capabilities. The code will be made publicly available to facilitate further research.

Nikita Lad, Iva Grabarić Andonovski, Dana Compton et al.

Background: To enlist publishers and journals in promoting the UN Sustainable Development Goals (SDGs), the United Nations and the International Publishers Association (IPA) launched the SDG Publishers Compact in 2020, and the European Association of Science Editors (EASE) published its Environmental Manifesto, a set of recommendations for journal editors on how to contribute to reducing a journal’s carbon footprint. It is important to monitor the impact of these initiatives on journal policies for developing future recommendations.Objectives: The EASE and the Higher Education Sustainability Initiative (HESI) SDG Publishers Compact Fellows developed a survey to assess the progress made by signatories to the SDG Publishers Compact, detect obstacles that prevent other publishers or journals from signing the compact, assess awareness and implementation of the EASE Environmental Manifesto, and identify other initiatives that promote SDGs.Methods: A multi-stakeholder group was formed, which included editors and both commercial and non-profit publishers, to design questions suited to journals and organizations at different stages of sustainability action. The survey was designed using SurveyMonkey, introduced in an online workshop, distributed through mail-ing lists to more than 2000 addresses, and promoted on social networks, and a total of 79 responses were collected and discussed.Results: Most respondents were representatives of smaller journals based in Europe. The majority were aware of the SDGs, but only half were aware of the SDG Publishers Compact, and only 17 (22%) were signatories to the Compact. Lack of awareness was the major reason for not joining the initiative, followed by lack of time or resources. Respondents focused mostly on quality education, and the majority were acting to achieve at least one SDG. Signatories to the compact mostly have a written environmental policy, have appointed an environmental officer, and are acquiring content related to the SDGs and promoting related activities. Non-signatories are also acting to minimize their environmental impact but have not considered the SDGs in their workflows. Both groups mainly do not have a dedicated budget to achieve the SDGs and have not completed a baseline of their activities. Activities undertaken to reach the SDGs had the most effect on community awareness. Half the respondents were members of EASE and were taking actions aligned with the Environmental Manifesto, mostly towards reducing their journal’s carbon footprint, and 25% are fol-lowing other initiatives aimed at achieving the SDGs as well.Conclusions: The survey showed that editors of small academic journals were not aware of the SDG Publishers Compact, although most of them are acting to achieve at least one SDG. Signatories to the Compact are implementing SDGs into their work-flows and practices, which shows the importance of the initiative. Greater efforts should be undertaken to make the editors of smaller journals aware of the Compact, encourage them to become its signatories, and provide them with more resources and metrics for monitoring their activities.

Chamani Shiranthika, Hadi Hadizadeh, Parvaneh Saeedi et al.

Federated Learning (FL) enables multiple clients to train a collaborative model without sharing their local data. Split Learning (SL) allows a model to be trained in a split manner across different locations. Split-Federated (SplitFed) learning is a more recent approach that combines the strengths of FL and SL. SplitFed minimizes the computational burden of FL by balancing computation across clients and servers, while still preserving data privacy. This makes it an ideal learning framework across various domains, especially in healthcare, where data privacy is of utmost importance. However, SplitFed networks encounter numerous communication challenges, such as latency, bandwidth constraints, synchronization overhead, and a large amount of data that needs to be transferred during the learning process. In this paper, we propose SplitFedZip -- a novel method that employs learned compression to reduce data transfer in SplitFed learning. Through experiments on medical image segmentation, we show that learned compression can provide a significant data communication reduction in SplitFed learning, while maintaining the accuracy of the final trained model. The implementation is available at: \url{https://github.com/ChamaniS/SplitFedZip}.

Saad Abdul Ghani, Zizhao Wang, Peter Stone et al.

This paper introduces Dynamic Learning from Learned Hallucination (Dyna-LfLH), a self-supervised method for training motion planners to navigate environments with dense and dynamic obstacles. Classical planners struggle with dense, unpredictable obstacles due to limited computation, while learning-based planners face challenges in acquiring high-quality demonstrations for imitation learning or dealing with exploration inefficiencies in reinforcement learning. Building on Learning from Hallucination (LfH), which synthesizes training data from past successful navigation experiences in simpler environments, Dyna-LfLH incorporates dynamic obstacles by generating them through a learned latent distribution. This enables efficient and safe motion planner training. We evaluate Dyna-LfLH on a ground robot in both simulated and real environments, achieving up to a 25% improvement in success rate compared to baselines.

Yankai Jiang, Mingze Sun, Heng Guo et al.

Self-supervised learning (SSL) has recently achieved promising performance for 3D medical image analysis tasks. Most current methods follow existing SSL paradigm originally designed for photographic or natural images, which cannot explicitly and thoroughly exploit the intrinsic similar anatomical structures across varying medical images. This may in fact degrade the quality of learned deep representations by maximizing the similarity among features containing spatial misalignment information and different anatomical semantics. In this work, we propose a new self-supervised learning framework, namely Alice, that explicitly fulfills Anatomical invariance modeling and semantic alignment via elaborately combining discriminative and generative objectives. Alice introduces a new contrastive learning strategy which encourages the similarity between views that are diversely mined but with consistent high-level semantics, in order to learn invariant anatomical features. Moreover, we design a conditional anatomical feature alignment module to complement corrupted embeddings with globally matched semantics and inter-patch topology information, conditioned by the distribution of local image content, which permits to create better contrastive pairs. Our extensive quantitative experiments on three 3D medical image analysis tasks demonstrate and validate the performance superiority of Alice, surpassing the previous best SSL counterpart methods and showing promising ability for united representation learning. Codes are available at https://github.com/alibaba-damo-academy/alice.

Susana Perez Blazquez, Inas Hipolito

This paper argues that Machine Learning (ML) algorithms must be educated. ML-trained algorithms moral decisions are ubiquitous in human society. Sometimes reverting the societal advances governments, NGOs and civil society have achieved with great effort in the last decades or are yet on the path to be achieved. While their decisions have an incommensurable impact on human societies, these algorithms are within the least educated agents known (data incomplete, un-inclusive, or biased). ML algorithms are not something separate from our human idiosyncrasy but an enactment of our most implicit prejudices and biases. Some research is devoted to responsibility assignment as a strategy to tackle immoral AI behaviour. Yet this paper argues that the solution for AI ethical decision-making resides in algorithm education (as opposed to the training) of ML. Drawing from an analogy between ML and child education for social responsibility, the paper offers clear directions for responsible and sustainable AI design, specifically with respect to how to educate algorithms to decide ethically.

Sheng Yue, Guanbo Wang, Wei Shao et al.

This work aims to tackle a major challenge in offline Inverse Reinforcement Learning (IRL), namely the reward extrapolation error, where the learned reward function may fail to explain the task correctly and misguide the agent in unseen environments due to the intrinsic covariate shift. Leveraging both expert data and lower-quality diverse data, we devise a principled algorithm (namely CLARE) that solves offline IRL efficiently via integrating "conservatism" into a learned reward function and utilizing an estimated dynamics model. Our theoretical analysis provides an upper bound on the return gap between the learned policy and the expert policy, based on which we characterize the impact of covariate shift by examining subtle two-tier tradeoffs between the exploitation (on both expert and diverse data) and exploration (on the estimated dynamics model). We show that CLARE can provably alleviate the reward extrapolation error by striking the right exploitation-exploration balance therein. Extensive experiments corroborate the significant performance gains of CLARE over existing state-of-the-art algorithms on MuJoCo continuous control tasks (especially with a small offline dataset), and the learned reward is highly instructive for further learning.

Alfiani Athma Putri Rosyadi, Adi Slamet Kusumawardana

Library is one of the learning resources that can be used by students to find the information needed. However, based on the result of observations obtained the fact that students are not attracted to the library because they have difficulty in finding the book they are looking for. This is because the management of books in the school still uses manual book data collection. Based on these problems, IT-based library management is one of the solutions offered. JIBAS is a network of school information systems that are integrated into assisting school management. One application that is available at JIBAS is SIMTAKA which is an application that helps schools manage data and library activities. The activity was started with the JIBAS program training by the Community Service Team. After the training, the activity continued with mentoring activities. This assistance was carried out in schools by testing the use of the JIBAS program in library management. The next activity is the implementation of the JIBAS program where partners have used the program in three months. The last activity is the reflection. In this activity, there is an exchange of information and experience during library management using the JIBAS program for three months. Suggestions for the next activity can be developed JIBAS for school finance and management.

Zuliyati Zuliyati, Edwin Djoenaedi, Hutomo Rusdianto

The purpose of this activity is the empowerment of small industries, especially in the cover craftsmen and report cards “GiGa” as service partners. The method used is technology guidance through the application of appropriate technology, assistance in the use of multi-use press machines using hydraulic, e-commerce, financial administration and management assistance. The precise technology of a multipurpose automatic press machine with a hydraulic machine can be used to produce a wide range of quality cover products with an effective and efficient production system. The results of this activity are increasing the ability of independent and resilient partners and increasing the role of the micro industry in development, creating job opportunities, increasing income for cover and report card craftsmen “GiGa”  in particular and the country in general, thus realizing a better economy. The output of this activity is an automatic multi-use press machine using hydraulic, effective and efficient production processes, quality cover products.

Olivier Pourret, David William Hedding, Daniel Enrique Ibarra et al.

Background: Open access (OA) implies free and unrestricted access to and re-use of research articles. Recently, OA publishing has seen a new wave of interest, debate, and practices surrounding that mode of publishing.Objectives: To provide an overview of publication practices and to compare them among six countries across the world to stimulate further debate and to raise awareness about OA to facilitate decision-making on further development of OA practices in earth sciences.Methods: The number of OA articles, their distribution among the six countries, and top ten journals publishing OA articles were identified using two databases, namely Scopus and the Web of Science, based mainly on the data for 2018.Results: In 2018, only 24%–31% of the total number of articles indexed by either of the databases were OA articles. Six of the top ten earth sciences journals that publish OA articles were fully OA journals and four were hybrid journals. Fully OA journals were mostly published by emerging publishers and their article processing charges ranged from $1000 to $2200.Conclusions: The rise in OA publishing has potential implications for researchers and tends to shift article-processing charges from organizations to individuals. Until the earth sciences community decides to move away from journal-based criteria to evaluate researchers, it is likely that such high costs will continue to maintain financial inequities within this research community, especially to the disadvantage of researchers from the least developed countries. However, earth scientists, by opting for legal self- archiving of their publications, could help to promote equitable and sustainable access to, and wider dissemination of, their work.

Srinivasan Arunachalam, Yihui Quek, John Smolin

Learning an unknown $n$-qubit quantum state $ρ$ is a fundamental challenge in quantum computing. Information-theoretically, it is known that tomography requires exponential in $n$ many copies of $ρ$ to estimate it up to trace distance. Motivated by computational learning theory, Aaronson et al. introduced many (weaker) learning models: the PAC model of learning states (Proceedings of Royal Society A'07), shadow tomography (STOC'18) for learning "shadows" of a state, a model that also requires learners to be differentially private (STOC'19) and the online model of learning states (NeurIPS'18). In these models it was shown that an unknown state can be learned "approximately" using linear-in-$n$ many copies of rho. But is there any relationship between these models? In this paper we prove a sequence of (information-theoretic) implications from differentially-private PAC learning, to communication complexity, to online learning and then to quantum stability. Our main result generalizes the recent work of Bun, Livni and Moran (Journal of the ACM'21) who showed that finite Littlestone dimension (of Boolean-valued concept classes) implies PAC learnability in the (approximate) differentially private (DP) setting. We first consider their work in the real-valued setting and further extend their techniques to the setting of learning quantum states. Key to our results is our generic quantum online learner, Robust Standard Optimal Algorithm (RSOA), which is robust to adversarial imprecision. We then show information-theoretic implications between DP learning quantum states in the PAC model, learnability of quantum states in the one-way communication model, online learning of quantum states, quantum stability (which is our conceptual contribution), various combinatorial parameters and give further applications to gentle shadow tomography and noisy quantum state learning.

M Ghofar Ismail, Siti Khoiriyah

Abstract The purpose of this study was to determine the implementation of intellectual development programs for students in the dormitory of MAN 2 Boyolali. This research used a qualitative descriptive approach which was conducted in March-June 2017 at the MAN 2 Boyolali Dormitory. Collecting data using the method of observation, interviews and documentation. The results of this study can be concluded as follows: (1) The implementation of the intellectual development program is carried out for boarding students of MAN 2 Boyolali. (2) Types of intellectual development programs: Foreign language development through Muhadhoroh, FTMP, General Sciences. (3) Using modern constructivist intellectual development methods, CTL, GLC, in addition to modern methods equipped with modern media to keep up with the times and supporting infrastructure (4) For example, at FTMP, students are asked to find and present a short story, and make short stories , making articles, opinions, editing a work, then exposing it on the internet and looking for many who comment or comment, write poetry, fiction and non-fiction, what stands out is scientific writing, until students in the dormitory win the writing of high school level scientific papers in Boyolali district. So that students in the dormitory in carrying out scientific papers are not so difficult. The implementation of intellectual development programs for students in dormitories has a positive effect on MAN 2 Boyolali. Keywords : Intellectual Development Program, Students, Dormitories Abstrak Tujuan penelitian ini adalah untuk mengetahui pelaksanaan program pengembangan intelektual bagi siswa di asrama MAN 2 Boyolali. Penelitian ini menggunakan pendekatan deskriptif kualitatif yang dilaksanakan pada bulan maret-juni 2017 di Asrama MAN 2 Boyolali. Pengumpulan data menggunakan metode observasi, wawancara dan dokumentasi. Hasil penelitian ini dapat disimpulkan sebagai berikut: (1) Pelaksanaan program pengembangan intelektual dilaksanakan untuk siswa asrama MAN 2 Boyolali. (2) Jenis-jenis program pengembangan intelektual: Pengembangan Bahasa asing melalui muhadhoroh, FTMP, Ilmu Umum. (3) Menggunakan metode pengembangan intelektual secara modern konstruktivisme, CTL, GLC, selain metode modern dilengkapi dengan media yang modern demi mengikuti perkembangan zaman dan sarana prasarana yang mendukung (4) Misalkan pada FTMP, siswa disuruh mencari dan mempresentasikan suatu cerpen, dan membuat cerpen, membuat artikel, opini, mengedit suatu karya, lalu mengekspos di internet dan mencari banyak yang mengkomen atau mengelike, menulis puisi, fiksi dan non fiksi, yang menonjol adalah karya tulis ilmiah, hingga siswa di asrama menjuarai penulisan karya ilmiah tingkat SMA satu kabupaten Boyolali. Sehingga siswa di asrama dalam mengerjalan karya tulis ilmiah tak begitu kesulitan. Pelaksanaan program pengembangan intelektual bagi siswa di asrama berpengaruh positif untuk MAN 2 Boyolali. Kata kunci : Program Pengembangan Intelektual, Siswa, Asrama

Ahmad Fahri Yahya Ainuri

Abstract As Indonesian people, of course we are familiar with religious groups with transnational ideologies that are oriented towards replacing government systems with Islamic systems (Imamat / Khilafah) based on the Qur'an and Hadith. Actually there is nothing wrong with the group's vision because the khilahfah system is a product of ijtihad of the predecessor ulama and normatively does not contradict Islamic law. It's just that the effort to coerce to change the law which has become a collective agreement in a country can legally be said as an act of rebellion and the act is not constitutionally justified. To address this phenomenon, the writer wants to give an understanding that implicitly our country (Indonesia) has actually implemented laws that are in accordance with Islamic sharia because the Pancasila ideology which is used as a national and state paradigm is fully in line with the sharia maqashid as contained in the Koran 'and Hadith which fully aims to educate people to become human beings who are deified, humane, united, just manifested into a common life (social life). Keywords: Islamic maqashid, Epistemology, Pancasila Education. Abstrak Sebagai masyarakat Indonesia, tentu kita tidak asing dengan adanya kelompok beragama dengan ideologi transnasional yang berorientasi mengganti sistem pemerintahan dengan sistem Islam (Imamah/Khilafah) yang berlandaskan al-Qur’an dan Hadis. Sebenarnya tidak ada yang salah dengan visi kelompok tersebut karena sistem khilahfah merupakan produk ijtihad para ulama pendahulu dan secara normatif tidak bertentangan dengan syariat Islam. Hanya saja, usaha melakukan paksaan untuk merubah undang-undang yang sudah menjadi kesepakatan bersama dalam suatu negara secara yuridis bisa dikatakan sebagai tindakan pemberontakan dan tindakan tersebut tidak dibenarkan secara konstitusional. Untuk mensikapi fenomena tersebut penulis ingin memberikan pemahaman bahwa secara implisit negara kita (Indonesia) sebenarnya sudah menerapkan undang-undang yang sesuai dengan syari’at Islam karena ideologi pancasila yang dijadikan sebagai paradigma berbangsa dan bernegara sepenuhnya sejalan dengan maqashid syariah yang tertuang dalam al-Qur’an dan Hadis yang spenuhnya bertujuan untuk mendidik masyarakat menjadi manusia yang berketuhanan, berperikemanusiaan, bersatu, adil yang termanifestasi ke dalam kehidupan bersama (kehidupan sosial). Kata Kunci : maqashid syariah, Epistemologi, Pendidikan Pancasila.

Niru Maheswaranathan, David Sussillo, Luke Metz et al.

Learned optimizers are algorithms that can themselves be trained to solve optimization problems. In contrast to baseline optimizers (such as momentum or Adam) that use simple update rules derived from theoretical principles, learned optimizers use flexible, high-dimensional, nonlinear parameterizations. Although this can lead to better performance in certain settings, their inner workings remain a mystery. How is a learned optimizer able to outperform a well tuned baseline? Has it learned a sophisticated combination of existing optimization techniques, or is it implementing completely new behavior? In this work, we address these questions by careful analysis and visualization of learned optimizers. We study learned optimizers trained from scratch on three disparate tasks, and discover that they have learned interpretable mechanisms, including: momentum, gradient clipping, learning rate schedules, and a new form of learning rate adaptation. Moreover, we show how the dynamics of learned optimizers enables these behaviors. Our results help elucidate the previously murky understanding of how learned optimizers work, and establish tools for interpreting future learned optimizers.

Kavosh Asadi, David Abel, Michael L. Littman

Can simple algorithms with a good representation solve challenging reinforcement learning problems? In this work, we answer this question in the affirmative, where we take "simple learning algorithm" to be tabular Q-Learning, the "good representations" to be a learned state abstraction, and "challenging problems" to be continuous control tasks. Our main contribution is a learning algorithm that abstracts a continuous state-space into a discrete one. We transfer this learned representation to unseen problems to enable effective learning. We provide theory showing that learned abstractions maintain a bounded value loss, and we report experiments showing that the abstractions empower tabular Q-Learning to learn efficiently in unseen tasks.

Firoj Alam, Shaden Shaar, Fahim Dalvi et al.

With the emergence of the COVID-19 pandemic, the political and the medical aspects of disinformation merged as the problem got elevated to a whole new level to become the first global infodemic. Fighting this infodemic has been declared one of the most important focus areas of the World Health Organization, with dangers ranging from promoting fake cures, rumors, and conspiracy theories to spreading xenophobia and panic. Addressing the issue requires solving a number of challenging problems such as identifying messages containing claims, determining their check-worthiness and factuality, and their potential to do harm as well as the nature of that harm, to mention just a few. To address this gap, we release a large dataset of 16K manually annotated tweets for fine-grained disinformation analysis that (i) focuses on COVID-19, (ii) combines the perspectives and the interests of journalists, fact-checkers, social media platforms, policy makers, and society, and (iii) covers Arabic, Bulgarian, Dutch, and English. Finally, we show strong evaluation results using pretrained Transformers, thus confirming the practical utility of the dataset in monolingual vs. multilingual, and single task vs. multitask settings.

Xinshi Chen, Hanjun Dai, Yu Li et al.

There is a recent surge of interest in designing deep architectures based on the update steps in traditional algorithms, or learning neural networks to improve and replace traditional algorithms. While traditional algorithms have certain stopping criteria for outputting results at different iterations, many algorithm-inspired deep models are restricted to a ``fixed-depth'' for all inputs. Similar to algorithms, the optimal depth of a deep architecture may be different for different input instances, either to avoid ``over-thinking'', or because we want to compute less for operations converged already. In this paper, we tackle this varying depth problem using a steerable architecture, where a feed-forward deep model and a variational stopping policy are learned together to sequentially determine the optimal number of layers for each input instance. Training such architecture is very challenging. We provide a variational Bayes perspective and design a novel and effective training procedure which decomposes the task into an oracle model learning stage and an imitation stage. Experimentally, we show that the learned deep model along with the stopping policy improves the performances on a diverse set of tasks, including learning sparse recovery, few-shot meta learning, and computer vision tasks.