Machine Learning under Resource Constraints - Fundamentals

Machine Learning under Resource Constraints explores novel algorithms that handle high-throughput data, high dimensions, or complex structures, focusing on minimizing resource consumption and executing predictions on diverse architectures. Volume 1 establishes the foundations, while Volume 2 covers application domains.

Format: Paperback / softback
Length: 505 pages
Publication date: 31 December 2022
Publisher: De Gruyter

Machine Learning under Resource Constraints delves into innovative machine learning algorithms that face the formidable challenge of processing high-throughput data, dealing with high dimensions, or navigating complex data structures. These algorithms are constrained by the interplay between the demands placed on data processing and the capabilities of the computing machinery. The resources at hand include runtime, memory, communication, and energy. Consequently, modern computer architectures assume a pivotal role in addressing these constraints. The goal of these algorithms is to optimize resource consumption while delivering accurate predictions. To achieve this, learned predictions are executed on diverse architectures, leveraging the potential of various computing systems to conserve resources.

Volume 1 serves as the foundational volume in this emerging field. It encompasses a comprehensive journey, from data collection and summarization to various aspects of resource-aware learning, including hardware, memory, energy, and communication awareness. Several machine learning methods are examined in terms of their resource requirements and strategies for enhancing their scalability across diverse computing architectures, spanning from embedded systems to large computing clusters.

Volume 2 explores the application of resource-constrained machine learning in real-world scenarios. It showcases case studies and practical examples from various domains, such as healthcare, finance, and social networks. The authors discuss how these algorithms have been developed and deployed to address resource-intensive problems and achieve significant improvements in efficiency and accuracy.

Volume 3 focuses on the theoretical aspects of resource-constrained machine learning. It delves into mathematical models, optimization techniques, and algorithms that can handle resource limitations effectively. The authors explore the trade-offs between different resource constraints and propose solutions to optimize the performance of machine learning systems.

Overall, Machine Learning under Resource Constraints provides a comprehensive overview of the novel approaches to machine learning research that consider resource constraints, as well as the application of the described methods in various domains of science and engineering. By exploring the challenges and opportunities posed by resource limitations, this trilogy empowers researchers and practitioners to develop more efficient and effective machine learning systems that can handle the vast amounts of data and complex structures prevalent in today's world.

Weight: 843g
Dimension: 240 x 170 (mm)
ISBN-13: 9783110785937