arXiv subjects

Fabio Porto

Publications and source records attributed to Fabio Porto.

At least 19 recordsLinked to original sources

arXiv2026

Gypscie: A Cross-Platform AI Artifact Management System

Artificial Intelligence (AI) models, encompassing both traditional machine learning (ML) and more advanced approaches such as deep learning and large language models (LLMs), play a central role in modern applications. AI model lifecycle management involves the end-to-end process of managing these models, from data collection and preparation to model building, evaluation, deployment, and continuous monitoring. This process is inherently complex, as it requires the coordination of diverse services that manage AI artifacts such as datasets, dataflows, and models, all orchestrated to operate seamlessly. In this context, it is essential to isolate applications from the complexity of interacting with heterogeneous services, datasets, and AI platforms. In this paper, we introduce Gypscie, a cross-platform AI artifact management system. By providing a unified view of all AI artifacts, the Gypscie platform simplifies the development and deployment of AI applications. This unified view is realized through a knowledge graph that captures application semantics and a rule-based query language that supports reasoning over data and models. Model lifecycle activities are represented as high-level dataflows that can be scheduled across multiple platforms, such as servers, cloud platforms, or supercomputers. Finally, Gypscie records provenance information about the artifacts it produces, thereby enabling explainability. Our qualitative comparison with representative AI systems shows that Gypscie supports a broader range of functionalities across the AI artifact lifecycle. Our experimental evaluation demonstrates that Gypscie can successfully optimize and schedule dataflows on AI platforms from an abstract specification.

cs.AI↗

arXiv2025

Towards a Spatiotemporal Fusion Approach to Precipitation Nowcasting

With the increasing availability of meteorological data from various sensors, numerical models and reanalysis products, the need for efficient data integration methods has become paramount for improving weather forecasts and hydrometeorological studies. In this work, we propose a data fusion approach for precipitation nowcasting by integrating data from meteorological and rain gauge stations in Rio de Janeiro metropolitan area with ERA5 reanalysis data and GFS numerical weather prediction. We employ the spatiotemporal deep learning architecture called STConvS2S, leveraging a structured dataset covering a 9 x 11 grid. The study spans from January 2011 to October 2024, and we evaluate the impact of integrating three surface station systems. Among the tested configurations, the fusion-based model achieves an F1-score of 0.2033 for forecasting heavy precipitation events (greater than 25 mm/h) at a one-hour lead time. Additionally, we present an ablation study to assess the contribution of each station network and propose a refined inference strategy for precipitation nowcasting, integrating the GFS numerical weather prediction (NWP) data with in-situ observations.

cs.LG↗

arXiv2025

Generative AI and the transformation of Work in Latin America -- Brazil

This survey explores the impact perceived by employers and employees of GenAI in their work activities in Brazil. Generative AI (GenAI) is gradually transforming Brazil workforce, particularly in micro and small businesses, though its adoption remains uneven. This survey examines the perceptions of employers and employees across five sectors: Sales, Customer Service, Graphic Design or Photography, Journalism or Content Production, and Software Development or Coding. The results are analyzed in light of six key dimensions of workforce impact. The findings reveal a mix of optimism, apprehension, and untapped potential in the integration of AI tools. This study serves as a foundation for developing inclusive strategies that maximize AI's benefits while safeguarding workers' rights. The IIA-LNCC supports open research and remains committed to shaping a future where technology and human potential progress together.

cs.CY↗

arXiv2024

StreamEnsemble: Predictive Queries over Spatiotemporal Streaming Data

Predictive queries over spatiotemporal (ST) stream data pose significant data processing and analysis challenges. ST data streams involve a set of time series whose data distributions may vary in space and time, exhibiting multiple distinct patterns. In this context, assuming a single machine learning model would adequately handle such variations is likely to lead to failure. To address this challenge, we propose StreamEnsemble, a novel approach to predictive queries over ST data that dynamically selects and allocates Machine Learning models according to the underlying time series distributions and model characteristics. Our experimental evaluation reveals that this method markedly outperforms traditional ensemble methods and single model approaches in terms of accuracy and time, demonstrating a significant reduction in prediction error of more than 10 times compared to traditional approaches.

stat.ML↗

arXiv2023

SoftED: Metrics for Soft Evaluation of Time Series Event Detection

Time series event detection methods are evaluated mainly by standard classification metrics that focus solely on detection accuracy. However, inaccuracy in detecting an event can often result from its preceding or delayed effects reflected in neighboring detections. These detections are valuable to trigger necessary actions or help mitigate unwelcome consequences. In this context, current metrics are insufficient and inadequate for the context of event detection. There is a demand for metrics that incorporate both the concept of time and temporal tolerance for neighboring detections. This paper introduces SoftED metrics, a new set of metrics designed for soft evaluating event detection methods. They enable the evaluation of both detection accuracy and the degree to which their detections represent events. They improved event detection evaluation by associating events and their representative detections, incorporating temporal tolerance in over 36\% of experiments compared to the usual classification metrics. SoftED metrics were validated by domain specialists that indicated their contribution to detection evaluation and method selection.

cs.LG↗

arXiv2022

Hyperspherical embedding for novel class classification

Deep learning models have become increasingly useful in many different industries. On the domain of image classification, convolutional neural networks proved the ability to learn robust features for the closed set problem, as shown in many different datasets, such as MNIST FASHIONMNIST, CIFAR10, CIFAR100, and IMAGENET. These approaches use deep neural networks with dense layers with softmax activation functions in order to learn features that can separate classes in a latent space. However, this traditional approach is not useful for identifying classes unseen on the training set, known as the open set problem. A similar problem occurs in scenarios involving learning on small data. To tackle both problems, few-shot learning has been proposed. In particular, metric learning learns features that obey constraints of a metric distance in the latent space in order to perform classification. However, while this approach proves to be useful for the open set problem, current implementation requires pair-wise training, where both positive and negative examples of similar images are presented during the training phase, which limits the applicability of these approaches in large data or large class scenarios given the combinatorial nature of the possible inputs.In this paper, we present a constraint-based approach applied to the representations in the latent space under the normalized softmax loss, proposed by[18]. We experimentally validate the proposed approach for the classification of unseen classes on different datasets using both metric learning and the normalized softmax loss, on disjoint and joint scenarios. Our results show that not only our proposed strategy can be efficiently trained on larger set of classes, as it does not require pairwise learning, but also present better classification results than the metric learning strategies surpassing its accuracy by a significant margin.

cs.CV↗

arXiv2021

A Simplicity Bubble Problem in Formal-Theoretic Learning Systems

When mining large datasets in order to predict new data, limitations of the principles behind statistical machine learning pose a serious challenge not only to the Big Data deluge, but also to the traditional assumptions that data generating processes are biased toward low algorithmic complexity. Even when one assumes an underlying algorithmic-informational bias toward simplicity in finite dataset generators, we show that current approaches to machine learning (including deep learning, or any formal-theoretic hybrid mix of top-down AI and statistical machine learning approaches), can always be deceived, naturally or artificially, by sufficiently large datasets. In particular, we demonstrate that, for every learning algorithm (with or without access to a formal theory), there is a sufficiently large dataset size above which the algorithmic probability of an unpredictable deceiver is an upper bound (up to a multiplicative constant that only depends on the learning algorithm) for the algorithmic probability of any other larger dataset. In other words, very large and complex datasets can deceive learning algorithms into a ``simplicity bubble'' as likely as any other particular non-deceiving dataset. These deceiving datasets guarantee that any prediction effected by the learning algorithm will unpredictably diverge from the high-algorithmic-complexity globally optimal solution while converging toward the low-algorithmic-complexity locally optimal solution, although the latter is deemed a global one by the learning algorithm. We discuss the framework and additional empirical conditions to be met in order to circumvent this deceptive phenomenon, moving away from statistical machine learning towards a stronger type of machine learning based on, and motivated by, the intrinsic power of algorithmic information theory and computability theory.

cs.IT↗

arXiv2021

Estimation of COVID-19 under-reporting in Brazilian States through SARI

Due to its impact, COVID-19 has been stressing the academy to search for curing, mitigating, or controlling it. However, when it comes to controlling, there are still few studies focused on under-reporting estimates. It is believed that under-reporting is a relevant factor in determining the actual mortality rate and, if not considered, can cause significant misinformation. Therefore, the objective of this work is to estimate the under-reporting of cases and deaths of COVID-19 in Brazilian states using data from the Infogripe on notification of Severe Acute Respiratory Infection (SARI). The methodology is based on the concepts of inertia and the use of event detection techniques to study the time series of hospitalized SARI cases. The estimate of real cases of the disease, called novelty, is calculated by comparing the difference in SARI cases in 2020 (after COVID-19) with the total expected cases in recent years (2016 to 2019) derived from a seasonal exponential moving average. The results show that under-reporting rates vary significantly between states and that there are no general patterns for states in the same region in Brazil. The published version of this paper is made available at https://doi.org/10.1007/s00354-021-00125-3. Please cite as: B. Paixão, L. Baroni, M. Pedroso, R. Salles, L. Escobar, C. de Sousa, R. de Freitas Saldanha, J. Soares, R. Coutinho, et al., 2021, Estimation of COVID-19 Under-Reporting in the Brazilian States Through SARI, New Generation Computing

stat.AP↗

arXiv2020

DJEnsemble: On the Selection of a Disjoint Ensemble of Deep Learning Black-Box Spatio-Temporal Models

In this paper, we present a cost-based approach for the automatic selection and allocation of a disjoint ensemble of black-box predictors to answer predictive spatio-temporal queries. Our approach is divided into two parts -- offline and online. During the offline part, we preprocess the predictive domain data -- transforming it into a regular grid -- and the black-box models -- computing their spatio-temporal learning function. In the online part, we compute a DJEnsemble plan which minimizes a multivariate cost function based on estimates for the prediction error and the execution cost -- producing a model spatial allocation matrix -- and run the optimal ensemble plan. We conduct a set of extensive experiments that evaluate the DJEnsemble approach and highlight its efficiency. We show that our cost model produces plans with performance close to the actual best plan. When compared against the traditional ensemble approach, DJEnsemble achieves up to $4X$ improvement in execution time and almost $9X$ improvement in prediction accuracy. To the best of our knowledge, this is the first work to solve the problem of optimizing the allocation of black-box models to answer predictive spatio-temporal queries.

cs.AI↗

arXiv2020

STConvS2S: Spatiotemporal Convolutional Sequence to Sequence Network for Weather Forecasting

Applying machine learning models to meteorological data brings many opportunities to the Geosciences field, such as predicting future weather conditions more accurately. In recent years, modeling meteorological data with deep neural networks has become a relevant area of investigation. These works apply either recurrent neural networks (RNN) or some hybrid approach mixing RNN and convolutional neural networks (CNN). In this work, we propose STConvS2S (Spatiotemporal Convolutional Sequence to Sequence Network), a deep learning architecture built for learning both spatial and temporal data dependencies using only convolutional layers. Our proposed architecture resolves two limitations of convolutional networks to predict sequences using historical data: (1) they violate the temporal order during the learning process and (2) they require the lengths of the input and output sequences to be equal. Computational experiments using air temperature and rainfall data from South America show that our architecture captures spatiotemporal context and that it outperforms or matches the results of state-of-the-art architectures for forecasting tasks. In particular, one of the variants of our proposed architecture is 23% better at predicting future sequences and five times faster at training than the RNN-based model used as a baseline.

cs.LG↗

arXiv2019

SAVIME: A Multidimensional System for the Analysis and Visualization of Simulation Data

Scientific applications produce a huge amount of data, which imposes serious management and analysis challenges. In particular, limitations in current database management systems prevent their adoption in simulation applications, in which in-situ analysis libraries, in-transit I/O interfaces and scientific format files are preferred over DBMSs. In order to make simulation applications benefit from DBMS support, the author proposes the development of a system called SAVIME in the context of his PhD thesis. SAVIME is an array database system designed to manage numerical simulation data. In this document, the author presents all work conducted so far and the current state of development.

cs.DB↗

arXiv2018

Towards In-transit Analysis on Supercomputing Environments

The drive towards exascale computing is opening an enormous opportunity for more realistic and precise simulations of natural phenomena. The process of simulation, however, involves not only the numerical computation of predictions but also the analysis of results both to evaluate the simulation quality and interpret the simulated phenomenon. In this context, one may consider the duality between transaction and analytical processing to be repositioned in this new context. The co-habitation of simulation computation and analysis has been named after in situ analysis, whereas the separation in different systems considered as in-transit analysis. In this paper we focus in the latter model and study the impact of transferring varying block size data from the simulation system to the analytical one. We use the Remote Direct Memory Access protocol (RDMA) that reduces the interference on performance caused by data copies and context switching. It adopts an in-memory data transfer strategy combined with TCP, using the BSD sockets API and the Linux splice(2) syscall. We present a performance evaluation with our work and traditional utilities.

cs.DC↗

arXiv2018

Parallel Computation of PDFs on Big Spatial Data Using Spark

We consider big spatial data, which is typically produced in scientific areas such as geological or seismic interpretation. The spatial data can be produced by observation (e.g. using sensors or soil instrument) or numerical simulation programs and correspond to points that represent a 3D soil cube area. However, errors in signal processing and modeling create some uncertainty, and thus a lack of accuracy in identifying geological or seismic phenomenons. Such uncertainty must be carefully analyzed. To analyze uncertainty, the main solution is to compute a Probability Density Function (PDF) of each point in the spatial cube area. However, computing PDFs on big spatial data can be very time consuming (from several hours to even months on a parallel computer). In this paper, we propose a new solution to efficiently compute such PDFs in parallel using Spark, with three methods: data grouping, machine learning prediction and sampling. We evaluate our solution by extensive experiments on different computer clusters using big data ranging from hundreds of GB to several TB. The experimental results show that our solution scales up very well and can reduce the execution time by a factor of 33 (in the order of seconds or minutes) compared with a baseline method.

cs.DC↗

arXiv2017

Constellation Queries over Big Data

A geometrical pattern is a set of points with all pairwise distances (or, more generally, relative distances) specified. Finding matches to such patterns has applications to spatial data in seismic, astronomical, and transportation contexts. For example, a particularly interesting geometric pattern in astronomy is the Einstein cross, which is an astronomical phenomenon in which a single quasar is observed as four distinct sky objects (due to gravitational lensing) when captured by earth telescopes. Finding such crosses, as well as other geometric patterns, is a challenging problem as the potential number of sets of elements that compose shapes is exponentially large in the size of the dataset and the pattern. In this paper, we denote geometric patterns as constellation queries and propose algorithms to find them in large data applications. Our methods combine quadtrees, matrix multiplication, and unindexed join processing to discover sets of points that match a geometric pattern within some additive factor on the pairwise distances. Our distributed experiments show that the choice of composition algorithm (matrix multiplication or nested loops) depends on the freedom introduced in the query geometry through the distance additive factor. Three clearly identified blocks of threshold values guide the choice of the best composition algorithm. Finally, solving the problem for relative distances requires a novel continuous-to-discrete transformation. To the best of our knowledge this paper is the first to investigate constellation queries at scale.

cs.DB↗

arXiv2016

A note on the complexity of the causal ordering problem

In this note we provide a concise report on the complexity of the causal ordering problem, originally introduced by Simon to reason about causal dependencies implicit in systems of mathematical equations. We show that Simon's classical algorithm to infer causal ordering is NP-Hard---an intractability previously guessed but never proven. We present then a detailed account based on Nayak's suggested algorithmic solution (the best available), which is dominated by computing transitive closure---bounded in time by $O(|\mathcal V|\cdot |\mathcal S|)$, where $\mathcal S(\mathcal E, \mathcal V)$ is the input system structure composed of a set $\mathcal E$ of equations over a set $\mathcal V$ of variables with number of variable appearances (density) $|\mathcal S|$. We also comment on the potential of causal ordering for emerging applications in large-scale hypothesis management and analytics.

cs.AI↗

arXiv2015

Managing large-scale scientific hypotheses as uncertain and probabilistic data with support for predictive analytics

The sheer scale of high-resolution raw data generated by simulation has motivated non-conventional approaches for data exploration referred as `immersive' and `in situ' query processing of the raw simulation data. Another step towards supporting scientific progress is to enable data-driven hypothesis management and predictive analytics out of simulation results. We present a synthesis method and tool for encoding and managing competing hypotheses as uncertain data in a probabilistic database that can be conditioned in the presence of observations.

cs.DB↗

arXiv2014

$Υ$-DB: A system for data-driven hypothesis management and analytics

The vision of $Υ$-DB introduces deterministic scientific hypotheses as a kind of uncertain and probabilistic data, and opens some key technical challenges for enabling data-driven hypothesis management and analytics. The $Υ$-DB system addresses those challenges throughout a design-by-synthesis pipeline that defines its architecture. It processes hypotheses from their XML-based extraction to encoding as uncertain and probabilistic U-relational data, and eventually to their conditioning in the presence of observations. In this demo we present a first prototype of the $Υ$-DB system. We showcase its core innovative features by means of use case scenarios in computational science in which the hypotheses are extracted from a model repository on the web and evaluated (rated/ranked) as probabilistic data.

cs.DB↗

arXiv2014

Design-theoretic encoding of deterministic hypotheses as constraints and correlations into U-relational databases

In view of the paradigm shift that makes science ever more data-driven, in this paper we consider deterministic scientific hypotheses as uncertain data. In the form of mathematical equations, hypotheses symmetrically relate aspects of the studied phenomena. For computing predictions, however, deterministic hypotheses are used asymmetrically as functions. We refer to Simon's notion of structural equations in order to extract the (so-called) causal ordering embedded in a hypothesis. Then we encode it into a set of functional dependencies (fd's) that is basic input to a design-theoretic method for the synthesis of U-relational databases (DB's). The causal ordering captured from a formally-specified system of mathematical equations into fd's determines not only the constraints (structure), but also the correlations (uncertainty chaining) hidden in the hypothesis predictive data. We show how to process it effectively through original algorithms for encoding and reasoning on the given hypotheses as constraints and correlations into U-relational DB's. The method is applicable to both quantitative and qualitative hypotheses and has underwent initial tests in a realistic use case from computational science.

cs.DB↗