Article
Computer Science, Information Systems
Xuekai Wei, Mingliang Zhou, Sam Kwong, Hui Yuan, Shiqi Wang, Guopu Zhu, Jingchao Cao
Summary: This paper proposes a reinforcement learning-based DASH technique to address user quality of experience, formulating the DASH adaptive bitrate selection problem as a Markov decision process. Experimental results show that the proposed RL-based ABR algorithm outperforms existing schemes in terms of both temporal and visual QoE factors.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Information Systems
Qixuan Zhang, Xinyi Weng, Guangyou Zhou, Yi Zhang, Jimmy Xiangji Huang
Summary: This paper proposes a new adaptive reinforcement learning framework for complex knowledge base question answering, and empirical results show that it is more effective than existing methods.
INFORMATION PROCESSING & MANAGEMENT
(2022)
Article
Computer Science, Artificial Intelligence
Weifeng Gao, Qianlong Dang, Maoguo Gong
Summary: This paper proposes an adaptive framework, STCS, for selecting coordinate systems in evolutionary algorithms. By constructing eigen coordinate systems using an archive-based covariance matrix, the issue of ineffective matching of different function landscapes is addressed. Experimental results demonstrate the efficiency and competitiveness of STCS across multiple test suites.
APPLIED SOFT COMPUTING
(2022)
Article
Computer Science, Information Systems
Tong Zhang, Yu Gou, Jun Liu, Tingting Yang, Jun-Hong Cui
Summary: The article introduces an underwater distributed and adaptive resource management framework (UDARMF) that maximizes network capacity by supporting more communications and resolves the contradiction between maximizing local capacity and global concurrency. The experimental results demonstrate the advantages of this framework in terms of network capacity and concurrency, considering different communication requirements and energy storage.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Chemistry, Multidisciplinary
Jeongho Kang, Kwangsue Chung
Summary: In this study, an adaptive streaming scheme using online reinforcement learning is proposed to improve the quality of video streaming. The scheme adapts to changes in client environment by upgrading the ABR model and utilizes state-of-the-art reinforcement learning algorithm to train the neural network model during video streaming.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Artificial Intelligence
Marcel Panzer, Norbert Gronau
Summary: In today's rapidly changing production landscape, flexible and resilient production process design is crucial for a company's competitiveness. This paper proposes a flexible framework that utilizes deep learning algorithms to optimize production processes, demonstrating high levels of control adaptability and multi-objective optimization performance.
JOURNAL OF INTELLIGENT MANUFACTURING
(2023)
Article
Computer Science, Hardware & Architecture
Min Qian, Yan-Fu Li
Summary: This paper proposes a novel adaptive undersampling framework that models the training process as a Markov decision process (MDP) to dynamically generate effective and divergent subsets. The framework is optimized using the soft actor-critic reinforcement learning method. Experimental results show that the proposed framework outperforms 16 benchmark methods in terms of classification performance and robustness.
IEEE TRANSACTIONS ON RELIABILITY
(2023)
Article
Engineering, Electrical & Electronic
Constantine A. Kyriakopoulos, Ilias Gialampoukidis, Stefanos Vrochidis, Ioannis Kompatsiaris
Summary: Production lines in manufacturing can benefit from quality diagnosis methods that use learning techniques to adapt to runtime conditions, improve performance, and solve complex computational problems in real-time. This study presents an innovative framework that uses learning automata to detect early divergence of a product's physical parameters from a specified quality range. The framework reduces complexity and achieves near-optimal results, making it suitable for practical deployment.
Article
Computer Science, Artificial Intelligence
Qin Zhao, Yu Ding, Chen Lu, Chao Wang, Liang Ma, Laifa Tao, Jian Ma
Summary: The authors propose a fault diagnosis framework based on contrastive augmented deep reinforcement learning, which includes two-stage training with contrastive loss and adaptive reward function. Case studies on two public datasets demonstrate that the pretraining stage can provide a well-trained feature extraction model, leading to better fault diagnosis performance compared to other advanced models.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Engineering, Mechanical
Zhiying Qiu, Wu Wei, Xiongding Liu
Summary: This paper focuses on adaptive gait generation with reinforcement learning for a hexapod robot and proposes a hierarchical and modular framework. Experimental results show that superior reinforcement learning algorithms can converge in the framework, and the trained gait policy exhibits stability and lower transportation cost on flat terrain.
Article
Computer Science, Artificial Intelligence
Marco Mussi, Davide Lombarda, Alberto Maria Metelli, Francesco Trovo, Marcello Restelli
Summary: Automated Reinforcement Learning (AutoRL) is a research area that is gaining attention by addressing the challenges of traditional Reinforcement Learning (RL) techniques, such as data collection, algorithm selection, and hyper-parameter tuning. In this work, a general and flexible framework called ARLO is proposed to construct automated pipelines for AutoRL, with separate pipelines for offline and online RL settings. A Python implementation of these pipelines is provided as an open-source library, which has been tested on various domains and shown promising performance with limited human intervention.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Wen Qi, Haoyu Fan, Hamid Reza Karimi, Hang Su
Summary: This paper presents an adaptive reinforcement learning (RL) based multimodal data fusion (AdaRL-MDF) framework for teaching the robot hand to play Rock-Paper-Scissors (RPS) game with humans. The framework includes an adaptive learning mechanism to update the ensemble classifier, an RL model providing intellectual wisdom to the robot, and a multimodal data fusion structure offering resistance to interference. The corresponding experiments prove the mentioned functions of the AdaRL-MDF model. The comparison accuracy and computational time show the high performance of the ensemble model by combining k-nearest neighbor (k-NN) and deep convolutional neural network (DCNN). In addition, the depth vision-based k-NN classifier obtains a 100% identification accuracy so that the predicted gestures can be regarded as the real value. The demonstration illustrates the real possibility of HRC application. The theory involved in this model provides the possibility of developing HRC intelligence.
Article
Computer Science, Artificial Intelligence
Antoine Theberge, Christian Desrosiers, Maxime Descoteaux, Pierre-Marc Jodoin
Summary: Diffusion MRI tractography is the only non-invasive tool to assess the white-matter structural connectivity of a brain. Using deep reinforcement learning to address tractography issues has shown competitive results and stable performance when generalizing to new data.
MEDICAL IMAGE ANALYSIS
(2021)
Article
Telecommunications
Syed Khurram Mahmud, Yuanwei Liu, Yue Chen, Kok Keong Chai
Summary: An adaptive reinforcement learning framework is proposed to maximize the sum-rate for NOMA-UAV networks, showing superior performance compared to non-adaptive counterparts. The study also compares JMLD and SIC for their BER performance and implementation complexity, highlighting their differences in performance and complexity.
IEEE COMMUNICATIONS LETTERS
(2021)
Article
Computer Science, Information Systems
Samina Amin, M. Irfan Uddin, Ala Abdulsalam Alarood, Wali Khan Mashwani, Abdulrahman Alzahrani, Ahmed Omar Alzahrani
Summary: This paper proposes a Reinforcement Learning (RL) based smart e-learning framework with Markov Decision Process (MDP) to enhance the learning experience for each student by providing personalized and effective learning paths. Experimental findings show significant improvements and superiority over long-term sessions. These promising results offer a potential solution for further research.
Article
Computer Science, Information Systems
Ili Ko, Desmond Chambers, Enda Barrett
INTERNATIONAL JOURNAL OF INFORMATION SECURITY
(2020)
Article
Computer Science, Interdisciplinary Applications
Rachael Shaw, Enda Howley, Enda Barrett
SIMULATION MODELLING PRACTICE AND THEORY
(2020)
Article
Computer Science, Information Systems
M. S. Miah, M. Schukat, E. Barrett
COMPUTER COMMUNICATIONS
(2020)
Article
Telecommunications
Mohammad Amzad Hossain, Michael Schukat, Enda Barrett
Summary: This paper proposes the concept of multiple reporting channels to improve the reporting time delay in cluster-based CCRNs, by allocating individual reporting channels based on frequency division multiple access to extend the sensing time of SUs. This approach enhances the decision accuracy of the FC and reduces the reporting time delay of CHs in comparison to conventional approaches.
WIRELESS PERSONAL COMMUNICATIONS
(2021)
Article
Computer Science, Information Systems
Rachael Shaw, Enda Howley, Enda Barrett
Summary: This paper explores the application of reinforcement learning algorithms for the VM consolidation problem in order to optimize the distribution of virtual machines and improve resource management in data centers. The empirical results demonstrate a 25% improvement in energy efficiency and a 63% reduction in service violations compared to a popular heuristic algorithm.
INFORMATION SYSTEMS
(2022)
Article
Computer Science, Information Systems
Ili Ko, Desmond Chambers, Enda Barrett
JOURNAL OF INFORMATION SECURITY AND APPLICATIONS
(2020)
Article
Computer Science, Artificial Intelligence
Ili Ko, Desmond Chambers, Enda Barrett
Summary: The paper discusses the use of Deep Learning algorithms to improve DDoS mitigation systems, proposing an intelligent attack mitigation (IAM) system, and introduces an ensemble approach using Recurrent Autonomous Autoencoders.
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
(2021)
Article
Computer Science, Information Systems
Daniel Kelly, Frank G. Glavin, Enda Barrett
Summary: Serverless computing is a new paradigm in cloud computing, offering a powerful development framework but also giving rise to new forms of cyber-attacks. This paper defines and identifies the threat of Denial of Wallet and its potential attack patterns, as well as demonstrates how it can circumvent existing mitigation systems. Additionally, it includes simulated experiments and a test bed for further research.
JOURNAL OF INFORMATION SECURITY AND APPLICATIONS
(2021)
Review
Green & Sustainable Science & Technology
Mostafa Rezaeimozafar, Rory F. D. Monaghan, Enda Barrett, Maeve Duffy
Summary: The electric power industry is transitioning towards a carbon-free smart system, with the integration of renewable energy resources bringing new opportunities and challenges for system operators and end-users. Energy storage systems play a crucial role in maximizing these opportunities and mitigating potential challenges. This study focuses on BTM energy storage systems installed in end-users' premises and explores their potential capabilities and challenges in today's power system.
RENEWABLE & SUSTAINABLE ENERGY REVIEWS
(2022)
Proceedings Paper
Green & Sustainable Science & Technology
Mostafa Rezaeimozafar, Rory Monaghan, Enda Barrett, Maeve Duffy
Summary: This paper investigates the impact of electricity tariffs on residential photovoltaic systems and batteries, as well as the effects of COVID-influenced consumption patterns and increased subsidies for photovoltaic energy on battery scheduling. It proposes an optimal solution to the battery scheduling problem using a genetic algorithm to minimize electricity costs for customers.
2021 THE 9TH IEEE INTERNATIONAL CONFERENCE ON SMART ENERGY GRID ENGINEERING (SEGE 2021)
(2021)
Proceedings Paper
Engineering, Electrical & Electronic
Mohammad Amzad Hossain, Michael Schukat, Enda Barrett
Summary: This paper proposes a MU-MIMO antennas aided CB-CSS scheme for CR enabled IoV networks, which enhances sensing gain, sum rate and reduces global error probability compared to conventional SISO antenna based CSS and NCSS schemes.
2021 IEEE 93RD VEHICULAR TECHNOLOGY CONFERENCE (VTC2021-SPRING)
(2021)
Proceedings Paper
Computer Science, Hardware & Architecture
Daniel Kelly, Frank Glavin, Enda Barrett
2020 IEEE 13TH INTERNATIONAL CONFERENCE ON CLOUD COMPUTING (CLOUD 2020)
(2020)
Article
Computer Science, Theory & Methods
Ili Ko, Desmond Chambers, Enda Barrett
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2020)