@article{dimartinoExplainableAIClinical2022, title = {Explainable {{AI}} for Clinical and Remote Health Applications: A Survey on Tabular and Time Series Data}, shorttitle = {Explainable {{AI}} for Clinical and Remote Health Applications}, author = {Di Martino, Flavio and Delmastro, Franca}, year = {2022}, month = oct, journal = {Artificial Intelligence Review}, pages = {1--55}, issn = {0269-2821}, abstract = {Nowadays Artificial Intelligence (AI) has become a fundamental component of healthcare applications, both clinical and remote, but the best performing AI systems are often too complex to be self-explaining. Explainable AI (XAI) techniques are defined to unveil the reasoning behind the system's predictions and decisions, and they become even more critical when dealing with sensitive and personal health data. It is worth noting that XAI has not gathered the same attention across different research areas and data types, especially in healthcare. In particular, many clinical and remote health applications are based on tabular and time series data, respectively, and XAI is not commonly analysed on these data types, while computer vision and Natural Language Processing (NLP) are the reference applications. To provide an overview of XAI methods that are most suitable for tabular and time series data in the healthcare domain, this paper provides a review of the literature in the last 5 years, illustrating the type of generated explanations and the efforts provided to evaluate their relevance and quality. Specifically, we identify clinical validation, consistency assessment, objective and standardised quality evaluation, and human-centered quality assessment as key features to ensure effective explanations for the end users. Finally, we highlight the main research challenges in the field as well as the limitations of existing XAI methods.}, langid = {english}, pmcid = {PMC9607788}, pmid = {36320613}, keywords = {Clinical DSS,EHR,Explainable AI,Health,Remote patient monitoring,Time series}, annotation = {2 citations (Crossref) [2023-03-01]}, file = {/home/daniel/Zotero/storage/BI7PX4YW/Di Martino y Delmastro - 2022 - Explainable AI for clinical and remote health appl.pdf} } @article{dixitComparingAnalyzingApplications2021a, title = {Comparing and {{Analyzing Applications}} of {{Intelligent Techniques}} in {{Cyberattack Detection}}}, author = {Dixit, P. and Kohli, R. and {Acevedo-Duque}, A. and {Gonzalez-Diaz}, R.R. and Jhaveri, R.H.}, year = {2021}, journal = {Security and Communication Networks}, volume = {2021}, issn = {1939-0114}, abstract = {Now a day's advancement in technology increases the use of automation, mobility, smart devices, and application over the Internet that can create serious problems for protection and the privacy of digital data and raised the global security issues. Therefore, the necessity of intelligent systems or techniques can prevent and protect the data over the network. Cyberattack is the most prominent problem of cybersecurity and now a challenging area of research for scientists and researchers. These attacks may destroy data, system, and resources and sometimes may damage the whole network. Previously numerous traditional techniques were used for the detection and mitigation of cyberattack, but the techniques are not efficient for new attacks. Today's machine learning and metaheuristic techniques are popularly applied in different areas to achieve efficient computation and fast processing of complex data of the network. This paper is discussing the improvements and enhancement of security models, frameworks for the detection of cyberattacks, and prevention by using different machine learning and optimization techniques in the domain of cybersecurity. This paper is focused on the literature of different metaheuristic algorithms for optimal feature selection and machine learning techniques for the classification of attacks, and some of the prominent algorithms such as GA, evolutionary, PSO, machine learning, and others are discussed in detail. This study provides descriptions and tutorials that can be referred from various literature citations, references, or latest research papers. The techniques discussed are efficiently applied with high performance for detection, mitigation, and identification of cyberattacks and provide a security mechanism over the network. Hence, this survey presents the description of various existing intelligent techniques, attack datasets, different observations, and comparative studies in detail. {\copyright} 2021 Priyanka Dixit et al.}, langid = {english}, annotation = {10 citations (Crossref) [2023-06-12]}, file = {/home/daniel/Zotero/storage/DQISDTHS/Dixit et al. - 2021 - Comparing and Analyzing Applications of Intelligen.pdf;/home/daniel/Zotero/storage/N9F2BLMI/display.html} } @article{dobrevaMultimodalInstallationExploring2023, title = {A {{Multimodal Installation Exploring Gender Bias}} in {{Artificial Intelligence}}}, author = {Dobreva, M. and Rukavina, T. and Stamou, V. and Vidaki, A.N. and Zacharopoulou, L.}, year = {2023}, journal = {Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)}, volume = {14020 LNCS}, pages = {27--46}, issn = {0302-9743}, abstract = {The ``Blackbox AI'' installation, developed as part of the EthicAI = LABS project, seeks to raise awareness about the social impact and ethical dimension of artificial intelligence (AI). This interdisciplinary installation explores various domains to bring to light the underrepresentation of women in STEM fields and the biases present in AI applications. The gender-swapped stories of women's experiences of discrimination in the workplace, collected by survey, showcase common patterns and explore the effect of flipping the gender. The text-to-image generation experiment highlights a preference for men in STEM professions and the prevalence of social and racial biases. The facial recognition examples demonstrate the discriminatory effects of such technologies on women, while the image generation investigation poses questions about the influence of AI technology on beauty, with the aim to empower women by pointing out bias in AI tools. The ultimate goal of the project is to challenge visitors to rethink their role in creating our digital future and address the issue of gender bias in artificial intelligence. {\copyright} 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.}, isbn = {9783031356803}, langid = {english}, keywords = {artificial intelligence (AI),discrimination,gender bias,social impact,technology}, annotation = {0 citations (Crossref) [2023-09-18]}, file = {/home/daniel/Zotero/storage/MLHDPAZ9/Dobreva et al_2023_A Multimodal Installation Exploring Gender Bias in Artificial Intelligence.pdf} } @article{guptaSurveyFederatedLearningApproaches2022, title = {Survey on {{Federated-Learning Approaches}} in {{Distributed Environment}}}, author = {Gupta, R. and Alam, T.}, year = {2022}, journal = {Wireless Personal Communications}, volume = {125}, number = {2}, pages = {1631--1652}, issn = {0929-6212}, abstract = {Federated-Learning (FL), a new paradigm in the machine-learning approach, wherein the clients train the global model collaboratively across various computational distributed units. The participants of the FL-networks performs communication with the centralized server without the exchange of sample data. This mechanism permits the users to obtain the richer global model performing training upon the larger data points. In this study, various researches of federated learning in distributed environment have been analysed. The Federated-learning framework model is implemented in centralized, decentralized and heterogeneous approach. Further, the privacy of the data collaborations and maintenance of secured framework in FL is focused. Differential-privacy technique is highly concentrated in various researches as the standardized method for mitigating those privacy risks. In some FL models, such as DRL-Deep reinforcement learning model is evolved for assisting the edge computing in a distributed environment, are highly focused in various studies. FedGRU-algorithm for traffic-flow prediction, non-IID, non-balanced, sparse and distributed attributes of federated-optimization is also analysed. The federated-learning framework contributes to obtain the global model in distributed systems handling heterogeneous resources in certain researches. The latter section of the paper demonstrates the critical analysis of the study, and the parameters relying upon the federated learning model were analysed. {\copyright} 2022, The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.}, langid = {english}, keywords = {Block chain network,Centralized,Differential-privacy,Distributed system,Edge-computing systems,FL-Federated-learning,Heterogeneous federated-learning,IoT-Internet of things}, annotation = {6 citations (Crossref) [2023-06-15]}, file = {/home/daniel/Zotero/storage/DZWPB5X8/display.html} } @article{liSurveyEvolutionaryDeep2023, title = {Survey on {{Evolutionary Deep Learning}}: {{Principles}}, {{Algorithms}}, {{Applications}}, and {{Open Issues}}}, shorttitle = {Survey on {{Evolutionary Deep Learning}}}, author = {Li, Nan and Ma, Lianbo and Yu, Guo and Xue, Bing and Zhang, Mengjie and Jin, Yaochu}, year = {2023}, month = sep, journal = {ACM Comput. Surv.}, volume = {56}, number = {2}, pages = {41:1--41:34}, issn = {0360-0300}, urldate = {2024-09-05}, abstract = {Over recent years, there has been a rapid development of deep learning (DL) in both industry and academia fields. However, finding the optimal hyperparameters of a DL model often needs high computational cost and human expertise. To mitigate the above issue, evolutionary computation (EC) as a powerful heuristic search approach has shown significant merits in the automated design of DL models, so-called evolutionary deep learning (EDL). This article aims to analyze EDL from the perspective of automated machine learning (AutoML). Specifically, we first illuminate EDL from DL and EC and regard EDL as an optimization problem. According to the DL pipeline, we systematically introduce EDL methods ranging from data preparation, model generation, to model deployment with a new taxonomy (i.e., what and how to evolve/optimize), and focus on the discussions of solution representation and search paradigm in handling the optimization problem by EC. Finally, key applications, open issues, and potentially promising lines of future research are suggested. This survey has reviewed recent developments of EDL and offers insightful guidelines for the development of EDL.}, annotation = {27 citations (Crossref) [2024-09-05]}, file = {/home/daniel/Zotero/storage/C2D74IYM/Li et al_2023_Survey on Evolutionary Deep Learning.pdf} } @article{majidDeepReinforcementLearning2024, title = {Deep {{Reinforcement Learning Versus Evolution Strategies}}: {{A Comparative Survey}}}, shorttitle = {Deep {{Reinforcement Learning Versus Evolution Strategies}}}, author = {Majid, Amjad Yousef and Saaybi, Serge and {Francois-Lavet}, Vincent and Prasad, R. Venkatesha and Verhoeven, Chris}, year = {2024}, month = sep, journal = {IEEE Transactions on Neural Networks and Learning Systems}, volume = {35}, number = {9}, pages = {11939--11957}, issn = {2162-2388}, urldate = {2024-09-05}, abstract = {Deep reinforcement learning (DRL) and evolution strategies (ESs) have surpassed human-level control in many sequential decision-making problems, yet many open challenges still exist. To get insights into the strengths and weaknesses of DRL versus ESs, an analysis of their respective capabilities and limitations is provided. After presenting their fundamental concepts and algorithms, a comparison is provided on key aspects, such as scalability, exploration, adaptation to dynamic environments, and multiagent learning. Current research challenges are also discussed, including sample efficiency, exploration versus exploitation, dealing with sparse rewards, and learning to plan. Then, the benefits of hybrid algorithms that combine DRL and ESs are highlighted.}, keywords = {Deep learning,Deep reinforcement learning (DRL),Evolution (biology),evolution strategies (ESs),exploration,Games,meta-learning,multiagent,Optimization,parallelism,Q-learning,Robots,Scalability}, annotation = {18 citations (Crossref) [2024-09-05]}, file = {/home/daniel/Zotero/storage/HS7FJ4FT/Majid et al_2024_Deep Reinforcement Learning Versus Evolution Strategies.pdf;/home/daniel/Zotero/storage/II2I6A5Z/10114063.html} } @article{yangGeneralizedOutofDistributionDetection2024a, title = {Generalized {{Out-of-Distribution Detection}}: {{A Survey}}}, shorttitle = {Generalized {{Out-of-Distribution Detection}}}, author = {Yang, Jingkang and Zhou, Kaiyang and Li, Yixuan and Liu, Ziwei}, year = {2024}, month = jun, journal = {International Journal of Computer Vision}, issn = {1573-1405}, urldate = {2024-07-29}, abstract = {Out-of-distribution (OOD) detection is critical to ensuring the reliability and safety of machine learning systems. For instance, in autonomous driving, we would like the driving system to issue an alert and hand over the control to humans when it detects unusual scenes or objects that it has never seen during training time and cannot make a safe decision. The term, OOD detection, first emerged in 2017 and since then has received increasing attention from the research community, leading to a plethora of methods developed, ranging from classification-based to density-based to distance-based ones. Meanwhile, several other problems, including anomaly detection (AD), novelty detection (ND), open set recognition (OSR), and outlier detection (OD), are closely related to OOD detection in terms of motivation and methodology. Despite common goals, these topics develop in isolation, and their subtle differences in definition and problem setting often confuse readers and practitioners. In this survey, we first present a unified framework called generalized OOD detection, which encompasses the five aforementioned problems, i.e.,AD, ND, OSR, OOD detection, and OD. Under our framework, these five problems can be seen as special cases or sub-tasks, and are easier to distinguish. Despite comprehensive surveys of related fields, the summarization of OOD detection methods remains incomplete and requires further advancement. This paper specifically addresses the gap in recent technical developments in the field of OOD detection. It also provides a comprehensive discussion of representative methods from other sub-tasks and how they relate to and inspire the development of OOD detection methods. The survey concludes by identifying open challenges and potential research directions.}, langid = {english}, keywords = {AI safety,Computer vision,Model trustworthiness,Open set recognition,Out-of-distribution detection}, annotation = {4 citations (Crossref) [2024-07-29]}, file = {/home/daniel/Zotero/storage/3FBRNUCU/Yang et al_2024_Generalized Out-of-Distribution Detection.pdf} }