How to Do Things With Words

What is modal logic

In recent years, mobile devices are equipped with increasingly advanced sensing and computing capabilities. Coupled with advancements in Deep Learning (DL), this opens up countless possibilities for meaningful applications, e.g., for medical purposes and in vehicular networks. Traditional cloud-based Machine Learning (ML) approaches require the data to be centralized in a cloud server or data center. However, this results in critical issues related to unacceptable latency and communication inefficiency. To this end, Mobile Edge Computing (MEC) has been proposed to bring intelligence closer to the edge, where data is produced. However, conventional enabling technologies for ML at mobile edge networks still require personal data to be shared with external parties, e.g., edge servers. Recently, in light of increasingly stringent data privacy legislations and growing privacy concerns, the concept of Federated Learning (FL) has been introduced. In FL, end devices use their local data to train an ML model required by the server. The end devices then send the model updates rather than raw data to the server for aggregation. FL can serve as an enabling technology in mobile edge networks since it enables the collaborative training of an ML model and also enables DL for mobile edge network optimization. However, in a large-scale and complex mobile edge network, heterogeneous devices with varying constraints are involved. This raises challenges of communication costs, resource allocation, and privacy and security in the implementation of FL at scale. In this survey, we begin with an introduction to the background and fundamentals of FL. Then, we highlight the aforementioned challenges of FL implementation and review existing solutions. Furthermore, we present the applications of FL for mobile edge network optimization. Finally, we discuss the important challenges and future research directions in FL.

/pdf/federated-learning-in-mobile-edge-networks-a-comprehensive-5dgelc5m6h.pdf

Federated Learning in Mobile Edge Networks: A Comprehensive Survey

In recent years, mobile devices are equipped with increasingly advanced sensing and computing capabilities. Coupled with advancements in Deep Learning (DL), this opens up countless possibilities for meaningful applications. Traditional cloudbased Machine Learning (ML) approaches require the data to be centralized in a cloud server or data center. However, this results in critical issues related to unacceptable latency and communication inefficiency. To this end, Mobile Edge Computing (MEC) has been proposed to bring intelligence closer to the edge, where data is produced. However, conventional enabling technologies for ML at mobile edge networks still require personal data to be shared with external parties, e.g., edge servers. Recently, in light of increasingly stringent data privacy legislations and growing privacy concerns, the concept of Federated Learning (FL) has been introduced. In FL, end devices use their local data to train an ML model required by the server. The end devices then send the model updates rather than raw data to the server for aggregation. FL can serve as an enabling technology in mobile edge networks since it enables the collaborative training of an ML model and also enables DL for mobile edge network optimization. However, in a large-scale and complex mobile edge network, heterogeneous devices with varying constraints are involved. This raises challenges of communication costs, resource allocation, and privacy and security in the implementation of FL at scale. In this survey, we begin with an introduction to the background and fundamentals of FL. Then, we highlight the aforementioned challenges of FL implementation and review existing solutions. Furthermore, we present the applications of FL for mobile edge network optimization. Finally, we discuss the important challenges and future research directions in FL

The Ecological Approach to Visual Perception

The following is a transcript of one of the discussion sessions that took place during the Workshop on Games, Action and Social Software at the Lorentz Center in Leiden. The discussion theme was set by the workshop organizers: "Is logic useful for the analysis of protocols, and if so, how?" The theme has attracted the usual protagonists, plus a cognitive scientist and a specialist in computer security.

What is protocol analysis

We use the example of the introduction of the anti-smoking legislation to model the relationship between the cultural make-up, in terms of values, of societies and the acceptance of and compliance with norms. We present two agent-based simulations and discuss the challenge of modeling sanctions and their relation to values and culture.

No smoking here: values, norms and culture in multi-agent systems

EU Personal Data Protection in Policy and Practice

Although the protection of personal data is harmonized within the EU by Directive 95/46/EC and will be further harmonized by the General Data Protection Regulation (GDPR) in 2018, there are significant differences in the ways in which EU member states implemented the protection of privacy and personal data in national laws, policies, and practices. This paper presents the main findings of a research project that compares the protection of privacy and personal data in eight EU member states: France, Germany, the UK, Ireland, Romania, Italy, Sweden, and the Netherlands. The comparison focuses on five major themes: awareness and trust, government policies for personal data protection, the applicable laws and regulations, implementation of those laws and regulations, and supervision and enforcement. The comparison of privacy and data protection regimes across the EU shows some remarkable findings, revealing which countries are front runners and which countries are lagging behind on specific aspects. For instance, the roles of and interplay between governments, civil rights organizations, and data protections authorities vary from country to country. Furthermore, with regard to privacy and data protection there are differences in the intensity and scope of political debates, information campaigns, media attention, and public debate. New concepts like privacy impact assessments, privacy by design, data breach notifications and big data are on the agenda in some but not all countries. Significant differences exist in (the levels of) enforcement by the different data protection authorities, due to different legal competencies, available budgets and personnel, policies, and cultural factors.

A Comparison of Data Protection Legislation and Policies Across the EU

In this paper, we present a prenex form theorem for a version of Independence Friendly logic, a logic with imperfect information. Lifting classical results to such logics turns out not to be straightforward, because independence conditions make the formulas sensitive to signalling phenomena. In particular, nested quantification over the same variable is shown to cause problems. For instance, renaming of bound variables may change the interpretations of a formula, there are only restricted quantifier extraction theorems, and slashed connectives cannot be so easily removed. Thus we correct some claims from Hintikka [8], Caicedo & Krynicki [3] and Hodges [11]. We refine definitions, in particular the notion of equivalence, and sharpen preconditions, allowing us to restore (restricted versions of) those claims, including the prenex form theorem of Caicedo & Krynicki [3], and, as a side result, we obtain an application to Skolem forms of classical formulas. It is a known fact that a complete calculus for IF-logic is impossible, but with our results we establish several quantifier rules that form a partial calculus of equivalence for a general version of IF-logic reflecting general properties of information flow in games.

Francien Dechesne

Papers

What is protocol analysis

No smoking here: values, norms and culture in multi-agent systems

EU Personal Data Protection in Policy and Practice

A Comparison of Data Protection Legislation and Policies Across the EU

Equivalence and quantifier rules for logic with imperfect information