Distribution shifts are all too common in real-world applications of machine learning. Domain adaptation (DA) aims to address this by providing various frameworks for adapting models to the deployment data without using labels. However, the domain shift scenario raises a second more subtle challenge: the difficulty of performing hyperparameter optimisation (HPO) for these adaptation algorithms without access to a labelled validation set. The unclear validation protocol for DA has led to bad practices in the literature, such as performing HPO using the target test labels when, in real-world scenarios, they are not available. This has resulted in over-optimism about DA research progress compared to reality. In this paper, we analyse the state of DA when using good evaluation practice, by benchmarking a suite of candidate validation criteria and using them to assess popular adaptation algorithms. We show that there are challenges across all three branches of domain adaptation methodology including Unsupervised Domain Adaptation (UDA), Source-Free Domain Adaptation (SFDA), and Test Time Adaptation (TTA). While the results show that realistically achievable performance is often worse th
Many machine learning classification tasks involve imbalanced datasets, which are often subject to over-sampling techniques aimed at improving model performance. However, these techniques are prone to generating unrealistic or infeasible samples. Furthermore, they often function as black boxes, lacking interpretability in their procedures. This opacity makes it difficult to track their effectiveness and provide necessary adjustments, and they may ultimately fail to yield significant performance improvements. To bridge this gap, we introduce the Decision Predicate Graphs for Data Augmentation (DPG-da), a framework that extracts interpretable decision predicates from trained models to capture domain rules and enforce them during sample generation. This design ensures that over-sampled data remain diverse, constraint-satisfying, and interpretable. In experiments on synthetic and real-world benchmark datasets, DPG-da consistently improves classification performance over traditional over-sampling methods, while guaranteeing logical validity and offering clear, interpretable explanations of the over-sampled data.
Domain adaptation (DA) is the topical problem of adapting models from labelled source datasets so that they perform well on target datasets where only unlabelled or partially labelled data is available. Many methods have been proposed to address this problem through different ways to minimise the domain shift between source and target datasets. In this paper we take an orthogonal perspective and propose a framework to further enhance performance by meta-learning the initial conditions of existing DA algorithms. This is challenging compared to the more widely considered setting of few-shot meta-learning, due to the length of the computation graph involved. Therefore we propose an online shortest-path meta-learning framework that is both computationally tractable and practically effective for improving DA performance. We present variants for both multi-source unsupervised domain adaptation (MSDA), and semi-supervised domain adaptation (SSDA). Importantly, our approach is agnostic to the base adaptation algorithm, and can be applied to improve many techniques. Experimentally, we demonstrate improvements on classic (DANN) and recent (MCD and MME) techniques for MSDA and SSDA, and ultim
Optical flow models trained on high-quality data often degrade severely when confronted with real-world corruptions such as blur, noise, and compression artifacts. To overcome this limitation, we formulate Degradation-Aware Optical Flow, a new task targeting accurate dense correspondence estimation from real-world corrupted videos. Our key insight is that the intermediate representations of image restoration diffusion models are inherently corruption-aware but lack temporal awareness. To address this limitation, we lift the model to attend across adjacent frames via full spatio-temporal attention, and empirically demonstrate that the resulting features exhibit zero-shot correspondence capabilities. Based on this finding, we present DA-Flow, a hybrid architecture that fuses these diffusion features with convolutional features within an iterative refinement framework. DA-Flow substantially outperforms existing optical flow methods under severe degradation across multiple benchmarks.
The computational generation of poems is a complex task, which involves several sound, prosodic and rhythmic resources. In this work we present PROPOE 2, with the extension of structural and rhythmic possibilities compared to the original system, generating poems from metered sentences extracted from the prose of Brazilian literature, with multiple rhythmic assembly criteria. These advances allow for a more coherent exploration of rhythms and sound effects for the poem. Results of poems generated by the system are demonstrated, with variations in parameters to exemplify generation and evaluation using various criteria. A geração computacional de poemas é uma tarefa complexa, que envolve diversos recursos sonoros, prosódicos e rítmicos. Neste trabalho apresentamos PROPOE 2, com a ampliação de possibilidades estruturais e rítmicas em relação ao sistema original, gerando poemas a partir de sentenças metrificadas extraídas da prosa da literatura brasileira, com múltiplos critérios rítmicos de montagem. Esses avanços permitem uma exploração mais coerente de ritmos e efeitos sonoros para o poema. Resultados de poemas gerados pelo sistema são demonstrados, com variações de parâmetros para
The thought-provoking analogy between AI and electricity, made by computer scientist and entrepreneur Andrew Ng, summarizes the deep transformation that recent advances in Artificial Intelligence (AI) have triggered in the world. This chapter presents an overview of the ever-evolving landscape of AI, written in Portuguese. With no intent to exhaust the subject, we explore the AI applications that are redefining sectors of the economy, impacting society and humanity. We analyze the risks that may come along with rapid technological progress and future trends in AI, an area that is on the path to becoming a general-purpose technology, just like electricity, which revolutionized society in the 19th and 20th centuries. A provocativa comparação entre IA e eletricidade, feita pelo cientista da computação e empreendedor Andrew Ng, resume a profunda transformação que os recentes avanços em Inteligência Artificial (IA) têm desencadeado no mundo. Este capítulo apresenta uma visão geral pela paisagem em constante evolução da IA. Sem pretensões de exaurir o assunto, exploramos as aplicações que estão redefinindo setores da economia, impactando a sociedade e a humanidade. Analisamos os riscos q
We explore confidential computing in the context of CBDCs using Microsoft's CCF framework as an example. By developing an experiment and comparing different approaches and performance and security metrics, we seek to evaluate the effectiveness of confidential computing to improve the privacy, security, and performance of CBDCs. Preliminary results suggest that confidential computing could be a promising solution to the technological challenges faced by CBDCs. Furthermore, by implementing confidential computing in DLTs such as Hyperledger Besu and utilizing frameworks such as CCF, we increase transaction confidentiality and privacy while maintaining the scalability and interoperability required for a global digital financial system. In conclusion, confidential computing can significantly bolster CBDC development, fostering a secure, private, and efficient financial future. -- Exploramos o uso da computação confidencial no contexto das CBDCs utilizando o framework CCF da Microsoft como exemplo. Via desenvolvimento de experimentos e comparação de diferentes abordagens e métricas de desempenho e segurança, buscamos avaliar a eficácia da computação confidencial para melhorar a privacida
This article presents and formalizes an elementary multiplication method discovered independently by a 10-year-old student, Anthony Lima Dias. The method reorganizes digit interactions in base-10 multiplication into a structured sequence of partial sums, reducing cognitive load and allowing reliable mental or semi-written computation. We provide a full mathematical proof of correctness, a comparison with the classical algorithm, formal notation, and a detailed contextual account of the discovery. The method expands the known catalog of student-invented algorithms and raises questions about cognitive pathways in arithmetic learning. Keywords: multiplication methods, mathematics education, mental calculation, student-invented algorithms, alternative algorithms, arithmetic strategies. -- -- Este artigo apresenta e formaliza um metodo elementar de multiplicacao descoberto de forma independente por um estudante de 10 anos, Anthony Lima Dias. O metodo reorganiza as interacoes entre algarismos na multiplicacao em base 10 em uma sequencia estruturada de somas parciais, reduzindo a carga cognitiva e permitindo um calculo mental ou semi-escrito mais confiavel. Fornecemos uma demonstracao mat
Dynamic manipulation is a key capability for advancing robot performance, enabling skills such as tossing. While recent learning-based approaches have pushed the field forward, most methods still rely on manually designed action parameterizations, limiting their ability to produce the highly coordinated motions required in complex tasks. Motion planning can generate feasible trajectories, but the dynamics gap-stemming from control inaccuracies, contact uncertainties, and aerodynamic effects-often causes large deviations between planned and executed trajectories. In this work, we propose Dynamics-Aware Motion Manifold Primitives (DA-MMP), a motion generation framework for goal-conditioned dynamic manipulation, and instantiate it on a challenging real-world ring-tossing task. Our approach extends motion manifold primitives to variable-length trajectories through a compact parameterization and learns a high-quality manifold from a large-scale dataset of planned motions. Building on this manifold, a conditional flow matching model is trained in the latent space with a small set of real-world trials, enabling the generation of throwing trajectories that account for execution dynamics. E
A study of the Revista General de Informacion y Documentacion, from 2005 to 2022. The objective is aimed at qualifying the structure of the research field and assessing the trajectory of the thematic areas covered. Applying as methodology the analysis of co-words, the construction of bibliometric networks and the creation of scientific maps. 514 documents are extracted from the Web of Science (WoS) database. The keywords assigned by the authors of the documents are selected and divided into three subperiods: 2005-2010, 2011-2016 and 2017-2022. In the results, 1701 author keywords and 37 bibliometric networks are obtained. In the period 2005-2010, the structure of the research field is represented on the scientific map with very few central and specialized topics, considering an initial and underdeveloped organization. In the period 2011-2016, the structure of the research field is distributed on the scientific map with a more varied number of central and specialized topics, but still insufficient, considering an organization in the process of development. In the period 2017-2022, the structure of the research field is shown on the map with all kinds of family of topics (central, sp
Several years ago Thérien and Wilke exhibited a decidable characterization of the languages of words that are definable in FO2(<,+1). Their proof relies on three separate ingredients. The first one is the characterization of the languages that are definable in FO2(<) as those whose syntactic semigroup belongs to the variety DA. Then, this result is combined with a wreath product argument showing that being definable in FO2(<,+1) corresponds to having a syntactic semigroup in DA*D. Finally, proving that membership of a semigroup in DA*D is decidable requires a third ingredient: the "locality" of DA, a result proved by Almeida. In this note we present a new self-contained and simple proof that definability in FO2(<,+1) is decidable. We obtain the locality of DA as a corollary.
This report presents the results of an exploratory analysis of the work context of Community Health Agents and Endemic Disease Control Agents in Primary Health Care (PHC), with a particular focus on Health Campaigns. To understand this context, the study adopted the Socially Aware Design framework, which employs artifacts and techniques to examine problem domains in a comprehensive and sociotechnical manner. Methods such as the Stakeholder Identification Diagram, Evaluation Frame, and Semiotic Framework were applied to identify stakeholders, anticipate challenges, and elicit social and technical requirements for the solution. Personas and Scenarios were also used to illustrate the potential impacts of a solution on various stakeholders and their life contexts within health campaigns. This report presents the analysis method, its application, and results, discussing the study's findings to inform the development of medium-fidelity prototypes for a PHC health campaign management solution.
Galaxy clusters are important cosmological probes since their abundance and spatial distribution are directly linked to structure formation on large scales. The principal uncertainty source on the cosmological parameter constraints concerns the cluster mass estimation from mass proxies. In addition, future surveys will provide a large amount of data, requiring an improvement in the accuracy of other elements used in the construction of cluster likelihoods. Therefore, accurate modeling of the mass-observable relations and reducing the effect of different systematic errors are fundamental steps for the success of cluster cosmology. In this work, we briefly review the abundance of galaxy clusters and discuss many sources of uncertainty. Os aglomerados de galáxias são importantes sondas cosmológicas, já que a abundância e a distribuição espacial desses objetos estão diretamente ligadas à formação de estruturas em grandes escalas. A maior fonte de incerteza nas restrições de parâmetros cosmológicos é originária das estimativas das massas dos aglomerados a partir da relação massa-observável. Além disso, os próximos grandes levantamentos fornecerão uma grande quantidade de dados, requeren
Descreve-se neste trabalho uma proposta de currículo interdisciplinar para a formação de professores de ciências da natureza. O curso permite a obtenção de quatro diplomas: professor de ciências para o ensino fundamental (nomenclatura brasileira), professor de biologia, física e química para o ensino médio. O diploma de professor de ciências é obtido com a integralização de créditos oferecidos ao longo dos três primeiros anos do curso. Para cada ano subsequente é possível obter os diplomas de professor do ensino médio. Os componentes curriculares pertinentes às ciências da natureza são inteiramente interdisciplinares nos três primeiros anos. No quarto ano são oferecidas disciplinas específicas de biologia, física e química, para a respectiva formação de professor do ensino médio. An interdisciplinary curriculum for science teaching undergraduate course will be described. The curriculum allows four degrees according the Brazilian educational legislation: science teacher for the middle school, biology, chemistry and physics teacher for the high school. The science teacher degree is obtained by accomplishing the three initial years syllabus. For each subsequent year it will be possibl
We describe automatic procedures for the selection of DA white dwarfs in the Hamburg/ESO objective-prism survey (HES). For this purpose, and the selection of other stellar objects (e.g., metal-poor stars and carbon stars), a flexible, robust algorithm for detection of stellar absorption and emission lines in the digital spectra of the HES was developed. Broad band (U-B, B-V) and narrow band (Strömgren c_1) colours can be derived directly from HES spectra, with precisions of sigma(U-B)=0.092mag; sigma(B-V)=0.095mag; sigma(c_1)=0.15mag. We describe simulation techniques that allow to convert model or slit spectra to HES spectra. These simulated objective-prism spectra are used to determine quantitative selection criteria, and for the study of selection functions. We present an atlas of simulated HES spectra of DA and DB white dwarfs. Our current selection algorithm is tuned to yield maximum efficiency of the candidate sample (minimum contamination with non-DAs). DA candidates are selected in the B-V versus U-B and c_1 versus W_λ(Hbeta+Hgamma+Hdelta) parameter spaces. The contamination of the resulting sample with hot subdwarfs is expected to be as low as ~8%, while there is essential
Non-linear dynamics is not a usually covered topic in undergraduate physics courses. However, its importance within classical mechanics and the general theory of dynamical systems is unquestionable. In this work we show that this subject can be included in the schedule of an introductory classical mechanics course without the need to develop a robust theory of chaotic dynamics. To do this, we take as examples conservative non-linear oscillators subject to time-dependent periodic forces. By introducing the concept of stroboscopic maps we show that it is possible to visualize the appearance of chaos in these systems. We also address the example of the forced simple pendulum applying the same treatment. Finally, we briefly comment on the more general theory of chaos in conservative Hamiltonian systems.
In this work, we discuss the determination of the distance to the Large Magellanic Cloud (LMC) using the Leavitt Law, utilizing the public catalog of Classical Cepheid Variable stars from the observational project OGLE-IV (The Optical Gravitational Lensing Experiment Collection of Variable Stars), consisting of 4709 stars in the Large Magellanic Cloud. To determine the pulsation period of Cepheid Variable stars, we employ the computational algorithm \textit{Lomb-Scargle periodogram} modified for our data. Additionally, with the calculation of the period, we can derive a period-luminosity relation for Cepheid Variables in the Large Magellanic Cloud and, using an independent calibration distance, deduce their distance moduli. We also discuss some general theoretical concepts of the physical mechanism behind the oscillation of variable stars.
What could metacognition look like in simple physical terms? We define metacognition as having beliefs about beliefs, which can be articulated very simply using the language of statistical physics and Bayesian mechanics. We introduce a typology between cognitive and metacognitive particles and develop an example of a metacognitive particle. This can be generalized to provide examples of higher forms of metacognition: i.e. particles having beliefs about beliefs about beliefs and so forth. We conclude by saying that the typology of particles laid down in the target article seems promising, for seemingly enabling a physics of cognition that builds upon and refines the free energy principle, toward a physical description of entities that specifically possess higher forms of cognition.
We present TA-DA, a new software aimed at greatly simplify and improve the analysis of stellar photometric data in comparison with theoretical models, and allow the derivation of stellar parameters from multi-band photometry. Its flexibility allows one to address a number of such problems: from the interpolation of stellar models, or sets of stellar physical parameters in general, to the computation of synthetic photometry in arbitrary filters or units; from the analysis of observed color-magnitude diagrams, to a Bayesian derivation of stellar parameters (and extinction) based on multi-band data. TA-DA is available as a pre-compiled IDL widget-based application; its graphical user interface makes it considerably user-friendly. In this paper we describe the software and its functionalities.
The contact process is a non-equilibrium Hamiltonian model that, even in one dimension, lacks an exact solution and has been extensively studied via Monte Carlo simulations, both in steady-state and time-dependent scenarios. Although the effects of particle mobility/diffusion on criticality have been preliminarily investigated, they remain incompletely understood. In this work, we examine how the critical rate of the model varies with the probability of particle mobility. By analyzing different stochastic evolutions of the system, we employ two modern approaches: 1) Random Matrix Theory (RMT): By building on the success of RMT, particularly Wishart-like matrices, in studying statistical physics of systems with up-down symmetry via magnetization dynamics [R. da Silva, IJMPC 2022], we demonstrate its applicability to models with an absorbing state. 2) Optimized Temporal Power Laws: By using short-time dynamics, we optimize power laws derived from ensemble-averaged evolutions of the system. Both methods consistently reveal that the critical rate decays with mobility according to a simple Belehradek function. Additionally, a straightforward mean-field analysis supports the decay of the