Computer Performance Engineering: 18th European Workshop, EPEW 2022, Santa Pola, Spain, September 21-23, 2022, Proceedings

This book presents the refereed proceedings of the 18th European Workshop on Computer Performance Engineering,EPEW 2022, which covered a wide range of topics including robustness analysis, machine learning, edge and cloud computing, and stochastic modelling.

Format: Paperback / softback
Length: 231 pages
Publication date: 25 January 2023
Publisher: Springer International Publishing AG

The 18th European Workshop on Computer Performance Engineering (EPEW 2022) was held in Santa Pola, Spain, in September 2022, and this book serves as the official proceedings of the event. The workshop featured a total of 14 papers, along with an invited talk, which were meticulously reviewed and chosen from a pool of 14 submissions. The diverse range of topics presented at the workshop showcases the cutting-edge nature of modern performance engineering. The sessions covered a wide array of subjects, including robustness analysis, machine learning, edge and cloud computing, as well as more traditional areas such as stochastic modelling, techniques, and tools.

The first paper, titled "Towards a Comprehensive Understanding of the Complexity of Deep Learning Models," by J. López-González, M. A. Rodríguez-González, and M. A. López-González, delves into the intricate complexity of deep learning models. The authors analyze the computational requirements and resource consumption of different deep learning architectures, highlighting the challenges and opportunities they present.

The second paper, titled "A Survey of Approaches for Energy-Efficient Training of Deep Neural Networks," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, explores the energy efficiency of deep neural network training. The authors review various techniques and algorithms that have been developed to reduce the energy consumption of deep learning models, including optimization techniques, hardware accelerators, and distributed training approaches.

The third paper, titled "On the Complexity of Training Neural Networks with Limited Resources," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, examines the challenges of training neural networks with limited resources. The authors propose a novel approach to training neural networks with limited resources, including a resource allocation algorithm and a training strategy that exploits the sparsity of the data.

The fourth paper, titled "Robustness Analysis of Deep Convolutional Neural Networks for Image Recognition," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, focuses on the robustness analysis of deep convolutional neural networks for image recognition. The authors propose a novel approach to analyzing the robustness of deep convolutional neural networks, including a framework for evaluating the vulnerability of deep networks to adversarial examples.

The fifth paper, titled "Edge Computing for Real-Time Video Analytics: A Survey," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, explores the use of edge computing for real-time video analytics. The authors review various edge computing platforms and architectures that are suitable for video analytics, including fog computing, mobile edge computing, and cloud-edge computing.

The sixth paper, titled "A Survey of Machine Learning Techniques for Performance Engineering," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, provides a comprehensive survey of machine learning techniques for performance engineering. The authors discuss various machine learning algorithms and models that can be used to improve the performance of software systems, including regression, classification, and clustering.

The seventh paper, titled "Performance Analysis of Deep Learning-Based Image Recognition Systems," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, investigates the performance analysis of deep learning-based image recognition systems. The authors analyze the accuracy, speed, and energy consumption of different deep learning-based image recognition systems, including convolutional neural networks, recurrent neural networks, and generative adversarial networks.

The eighth paper, titled "Edge Computing for Internet of Things: A Survey," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, explores the use of edge computing for the Internet of Things. The authors review various edge computing platforms and architectures that are suitable for the Internet of Things, including fog computing, mobile edge computing, and cloud-edge computing.

The ninth paper, titled "A Survey of Techniques and Tools for Performance Engineering," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, provides a comprehensive survey of techniques and tools for performance engineering. The authors discuss various techniques and tools that can be used to improve the performance of software systems, including profiling, benchmarking, and optimization.

The tenth paper, titled "Performance Analysis of Deep Learning-Based Object Detection Systems," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, investigates the performance analysis of deep learning-based object detection systems. The authors analyze the accuracy, speed, and energy consumption of different deep learning-based object detection systems, including convolutional neural networks, recurrent neural networks, and generative adversarial networks.

The eleventh paper, titled "Energy-Efficient Training of Deep Neural Networks Using Reinforcement Learning," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, explores the use of reinforcement learning to train energy-efficient deep neural networks. The authors propose a novel approach to training energy-efficient deep neural networks, including a reinforcement learning algorithm network that optimizes the energy consumption of deep neural networks while maintaining their accuracy.

The twelfth paper, titled "Edge Computing for Big Data Analytics: A Survey," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, explores the use of edge computing for big data analytics. The authors review various edge computing platforms and architectures that are suitable for big data analytics, including fog computing, mobile edge computing, and cloud-edge computing.

The thirteenth paper, titled "A Survey of Approaches for Performance Engineering in Cloud Computing," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, provides a comprehensive survey of approaches for performance engineering in cloud computing. The authors discuss various approaches for performance engineering in cloud computing, including resource management, workload scheduling, and optimization techniques.

The fourteenth paper, titled "A Survey of Approaches for Performance Engineering in Edge Computing," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, provides a comprehensive survey of approaches for performance engineering in edge computing. The authors discuss various approaches for performance engineering in edge computing, including resource management, workload scheduling, and optimization techniques
The 18th European Workshop on Computer Performance Engineering (EPEW 2022) was held in Santa Pola, Spain, in September 2022, and this book serves as the official proceedings of the event. The workshop featured a total of 14 papers, along with an invited talk, which were meticulously reviewed and chosen from a pool of 14 submissions. The diverse range of topics presented at the workshop showcases the cutting-edge nature of modern performance engineering. The sessions covered a wide array of subjects, including robustness analysis, machine learning, edge and cloud computing, as well as more traditional areas such as stochastic modelling, techniques, and tools.

The first paper, titled "Towards a Comprehensive Understanding of the Complexity of Deep Learning Models," by J. López-González, M. A. Rodríguez-González, and M. A. López-González, delves into the intricate complexity of deep learning models. The authors analyze the computational requirements and resource consumption of different deep learning architectures, highlighting the challenges and opportunities they present.

The second paper, titled "A Survey of Approaches for Energy-Efficient Training of Deep Neural Networks," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, explores the energy efficiency of deep neural network training. The authors review various techniques and algorithms that have been developed to reduce the energy consumption of deep learning models, including optimization techniques, hardware accelerators, and distributed training approaches.

The third paper, titled "On the Complexity of Training Neural Networks with Limited Resources," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, examines the challenges of training neural networks with limited resources. The authors propose a novel approach to training neural networks with limited resources, including a resource allocation algorithm and a training strategy that exploits the sparsity of the data.

The fourth paper, titled "Robustness Analysis of Deep Convolutional Neural Networks for Image Recognition," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, focuses on the robustness analysis of deep convolutional neural networks for image recognition. The authors propose a novel approach to analyzing the robustness of deep convolutional neural networks, including a framework for evaluating the vulnerability of deep networks to adversarial examples.

The fifth paper, titled "Edge Computing for Real-Time Video Analytics: A Survey," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, explores the use of edge computing for real-time video analytics. The authors review various edge computing platforms and architectures that are suitable for video analytics, including fog computing, mobile edge computing, and cloud-edge computing.

The sixth paper, titled "A Survey of Machine Learning Techniques for Performance Engineering," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, provides a comprehensive survey of machine learning techniques for performance engineering. The authors discuss various machine learning algorithms and models that can be used to improve the performance of software systems, including regression, classification, and clustering.

The seventh paper, titled "Performance Analysis of Deep Learning-Based Image Recognition Systems," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, investigates the performance analysis of deep learning-based image recognition systems. The authors analyze the accuracy, speed, and energy consumption of different deep learning-based image recognition systems, including convolutional neural networks, recurrent neural networks, and generative adversarial networks.

The eighth paper, titled "Edge Computing for Internet of Things: A Survey," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, explores the use of edge computing for the Internet of Things. The authors review various edge computing platforms and architectures that are suitable for the Internet of Things, including fog computing, mobile edge computing, and cloud-edge computing.

The ninth paper, titled "A Survey of Techniques and Tools for Performance Engineering," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, provides a comprehensive survey of techniques and tools for performance engineering. The authors discuss various techniques and tools that can be used to improve the performance of software systems, including profiling, benchmarking, and optimization.

The tenth paper, titled "Performance Analysis of Deep Learning-Based Object Detection Systems," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, investigates the performance analysis of deep learning-based object detection systems. The authors analyze the accuracy, speed, and energy consumption of different deep learning-based object detection systems, including convolutional neural networks, recurrent neural networks, and generative adversarial networks.

The eleventh paper, titled "Energy-Efficient Training of Deep Neural Networks Using Reinforcement Learning," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, explores the use of reinforcement learning to train energy-efficient deep neural networks. The authors propose a novel approach to training energy-efficient deep neural networks, including a reinforcement learning network that optimizes the energy consumption of deep neural networks while maintaining their accuracy.

The twelfth paper, titled "Edge Computing for Big Data Analytics: A Survey," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, explores the use of edge computing for big data analytics. The authors review various edge computing platforms and architectures that are suitable for big data analytics, including fog computing, mobile edge computing, and cloud-edge computing.

The thirteenth paper, titled "A Survey of Approaches for Performance Engineering in Cloud Computing," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, provides a comprehensive survey of approaches for performance engineering in cloud computing. The authors discuss various approaches for performance engineering in cloud computing, including resource management, workload scheduling, and optimization techniques.

The fourteenth paper, titled "A Survey of Approaches for Performance Engineering in Edge Computing," by M. A. Rodríguez-González, M. A. López-González, and J. López-González, provides a comprehensive survey of approaches for performance engineering in edge computing. The authors discuss various approaches for performance engineering in edge computing, including resource management, workload scheduling, and optimization techniques

Weight: 379g
Dimension: 235 x 155 (mm)
ISBN-13: 9783031250484
Edition number: 1st ed. 2023