This paper engages in a comprehensive investigation concerning the application of Explainable Artificial Intelligence (xAI) within the context of deep learning and Artificial Intelligence, with a specific focus on its implications for cybersecurity. Firstly, the paper gives an overview of xAI techniques and their significance and benefits when applied in cybersecurity. Subsequently, the authors methodically delineate their systematic mapping study, which serves as an investigative tool for discerning the potential trajectory of the field. This strategic methodological framework lets one identify the future research directions and opportunities that underlie the integration of xAI within the realm of Deep Learning, Artificial Intelligence, and cybersecurity, which are described in-depth. Then, the paper brings together all the gathered insights from this extensive investigation and closes with final conclusions.

In this paper we propose the federated private local training algorithm (Fed-PLT) for federated learning, to overcome the challenges of (i) expensive communications and (ii) privacy preservation. We address (i) by allowing for both partial participation and local training, which significantly reduce the number of communication rounds between the central coordinator and computing agents. The algorithm matches the state of the art in the sense that the use of local training demonstrably does not impact accuracy. Additionally, agents have the flexibility to choose from various local training solvers, such as (stochastic) gradient descent and accelerated gradient descent. Further, we investigate how employing local training can enhance privacy, addressing point (ii). In particular, we derive differential privacy bounds and highlight their dependence on the number of local training epochs. We assess the effectiveness of the proposed algorithm by comparing it to alternative techniques, considering both theoretical analysis and numerical results from a classification task.

Despite the fact that machine learning has been applied in innumerable domains, its models have usually operated in a black box fashion, i.e. without revealing the rationale behind their decisions. For human users, insufficient model transparency may result in the lack of trust in the technology, effectively hindering its development and adoption. This pressing need for the society to understand the reasoning for the model’s decisions gave rise to the concept of explainable AI (xAI), which has been the subject of extensive research. Since its introduction, a number of techniques providing explainability have been proposed. Yet, there has been no consensus reached regarding how to measure and evaluate the performance of the explainability methods. So far, the state-of-the-art literature has proposed two directions for evaluating explainable AI–the technical one and the human-centered one. Although the literature deems the technical way to be the objective one, it still suggests supplementing the evaluation with the subjective, human-centered approach, which proves to be time-consuming and requires considerable effort. This paper highlights the need to enhance and improve the existing technical metrics, to quantify the explainability of ML models in an objective, automated and time and cost-effective way. The text presents the current state-of-the-art in xAI evaluation metrics, discussing both human-centered and computer-centered approaches. Its contribution is in its comprehensive discussion of the evaluation methods for explainable AI (xAI).

Hybrid Artificial Intelligence (HAI) algorithms are well adapted to the industrial environment since they require significantly less data than data-driven Artificial Intelligence (AI) algorithms, matching the industry constraints. However, HAI does not fully address the issue of trust (validity, transparency, explainability, and ethics) which must be tackled to achieve world-class HAI beneficial to humans individually, organisationally, and societally. This paper focuses on describing the methods used to collect and elicit trustworthiness requirements in the European ULTIMATE project. It includes requirements register and a trustworthiness glossary of terms.

Template to support the elicitation and prioritisation of trustworthiness requirements for a use case of the ULTIMATE project. Composed mainly from the Ethics Guidelines for Trustworthy AI (AI-HLEG) and the Ethically Aligned Design (IEEE) documents.

In this paper we consider online distributed learning problems. Online distributed learning refers to the process of training learning models on distributed data sources. In our setting a set of agents need to cooperatively train a learning model from streaming data. Differently from federated learning, the proposed approach does not rely on a central server but only on peer-to-peer communications among the agents. This approach is often used in scenarios where data cannot be moved to a centralized location due to privacy, security, or cost reasons. In order to overcome the absence of a central server, we propose a distributed algorithm that relies on a quantized, finite-time coordination protocol to aggregate the locally trained models. Furthermore, our algorithm allows for the use of stochastic gradients during local training. Stochastic gradients are computed using a randomly sampled subset of the local training data, which makes the proposed algorithm more efficient and scalable than traditional gradient descent. In our paper, we analyze the performance of the proposed algorithm in terms of the mean distance from the online solution. Finally, we present numerical results for a logistic regression task.

Different visualisation solutions exist in the literature that allow for advanced analysis of structured data and the development of interactive dashboards. For example, there are various tools for robotics applications and general-purpose data visualisation software solutions. However, these tools do not always allow one to address the advanced needs related to robotic data analytics. In particular, we argue that the existing solutions for robotic applications heavily focus on sensor data diagnosis and debugging. As a result, they do not provide flexible ways to allow the data scientist to integrate AI algorithms. Therefore, in this paper, we propose a novel toolkit architecture that leverages modern web technologies in the process of robotic data inspection, visualisation, and analysis.