Abstract
Federated learning (FL) is a collaborative, decentralized privacy-preserving method to attach the challenges of storing data and data privacy. Artificial intelligence, machine learning, smart devices, and deep learning have strongly marked the last years. Two challenges arose in data science as a result. First, the regulation protected the data by creating the General Data Protection Regulation, in which organizations are not allowed to keep or transfer data without the owner's authorization. Another challenge is the large volume of data generated in the era of big data, and keeping that data in one only server becomes increasingly tricky. Therefore, the data is allocated into different locations or generated by devices, creating the need to build models or perform calculations without transferring data to a single location. The new term FL emerged as a sub-area of machine learning that aims to solve the challenge of making distributed models with privacy considerations. This survey starts by describing relevant concepts, definitions, and methods, followed by an in-depth investigation of federated model evaluation. Finally, we discuss three promising applications for further research: anomaly detection, distributed data streams, and graph representation.This article is categorized under:Technologies > Machine LearningTechnologies > Artificial Intelligence

Abstract
Abstract Fishery analysis is critical in maintaining the long-term sustainability of species and the livelihoods of millions of people who depend on fishing for food and income. The fishing gear, or metier, is a key factor significantly impacting marine habitats, selectively targeting species and fish sizes. Analysis of commercial catches or landings by metier in fishery stock assessment and management is crucial, providing robust estimates of fishing efforts and their impact on marine ecosystems. In this paper, we focus on a unique data set from the Azores’ fishing data collection programs between 2010 and 2017, where little information on metiers is available and sparse throughout our timeline. Our main objective is to tackle the task of data set reconstruction, leveraging domain knowledge and machine learning methods to retrieve or associate metier-related information to each fish landing. We empirically validate the feasibility of this task using a diverse set of modeling approaches and demonstrate how it provides new insights into different fisheries’ behavior and the impact of metiers over time, which are essential for future fish population assessments, management, and conservation efforts.

Abstract
Os autores apresentam a Aprendizagem Baseada em Soluções Efetivas que deriva da Aprendizagem Baseada em Projeto, mas aplicada a problemas reais com objetivo de contruir soluções efetivas. A enfase é colocada na efetividade no pressuposto que incentiva a um maior envolvimento e comprometimento por parte dos alunos, assegurando um contexto que se pretende mais aliciante e próximo do que será a realidade profissional dos alunos. A efetividade é aferida pelas funcionalidades consideradas essenciais à plena utilização e resolução do problema, mas também pela viabilidade da aplicação ser efetivamente utilizada, sem que seja necessário a continuidade do envolvimento dos alunos. As evidências empíricas apontam um claro aumento da aquisição de competências, do número de aprovados e das classificações. Permitiu também definir um posicionamento estratégico de cooperação com a comunidade envolvente, em que todas as partes beneficiam (formandos, docentes, instituição de ensino, entidades locais e regionais e empregadores).

Abstract
This tool paper presents Caos: a methodology and a programming framework for computer-aided design of structural operational semantics for formal models. This framework includes a set of Scala libraries and a workflow to produce visual and interactive diagrams that animate and provide insights over the structure and the semantics of a given abstract model with operational rules. Caos follows an approach in which theoretical foundations and a practical tool are built together, as an alternative to foundations-first design (tool justifies theory) or tool-first design (foundations justify practice). The advantage of Caos is that the tool-under-development can immediately be used to automatically run numerous and sizeable examples in order to identify subtle mistakes, unexpected outcomes, and unforeseen limitations in the foundations-under-development, as early as possible. We share two success stories of Caos' methodology and framework in our own teaching and research context, where we analyse a simple while-language and a choreographic language, including their operational rules and the concurrent composition of such rules. We further discuss how others can include Caos in their own analysis and Scala tools.

Abstract
Decent Work Agenda consists of a comprehensive initiative for promoting safety at work and social protection. Over 20 years since its conceptual release, measuring the progress of its elements is still challenging even after the publication of the decent work indicators guideline by the International Labour Organization in 2012. To close this gap, we use a Directional Distance Function (DDF) to measure the efficiency of safe work environment, and propose a combined measure taking into consideration also the efficacy. To illustrate the application of DDF in a reality-based case, we conducted a longitudinal study in a multinational organization. Data were collected from 21 branches of the company over 4 years (2018-2021). In the period of analysis, 60% of the branches were efficient in average, composing an overall efficiency score of 0.91. Our results also indicated low dispersion between the efficiency scores year on year due to a small-scale interquartile range. Finally, the use of efficiency combined with efficacy resulted in a promising approach for managerial applications. This research presents some contributions. One is the novelty approach of measuring the efficiency of safe work environment using a DDF model in a real-world application. Another is the managerial benefits of identifying benchmarks, as well as revealing potential improvements as a mechanism to reduce decent work deficits. From a modeling perspective, our conclusions suggest caution in considering only efficiency to measure safe work environment due to its relative nature. Thus, further studies are recommended to explore the use of combined measures in the analysis of decent work.

Abstract
The Brazilian Power System is mainly composed of renewable generation from hydroelectric and wind. Hence, spot and forward electricity prices tend to represent the inherently stochastic nature of these resources, while risk management is a measure taken by agents, especially hydro power plants (HPPs) to hedge against deep financial losses. A HPP goal is to maximize its profit considering uncertainties in forward electricity prices, spot prices, and generation scaling factor (GSF) for years ahead. Therefore, the objective of this work is to simulate the real decision-making process of a HPP, where they need to have a perspective of the forward market and future spot price assessment to negotiate forward electricity contracts. To do so, the present work models the uncertainty in electricity forward prices via two-stage stochastic programming, assessing the benefits of the stochastic solution in comparison to the deterministic one. In addition, different risk aversion levels are assessed using conditional value at risk (CVaR). An important conclusion is that the results show that the greater the HPP risk aversion is, the greater the energy selling via electricity forward contracts. Moreover, the proposed model has benefits in comparison to a deterministic approach.