Название: Explainable AI (XAI) for Sustainable Development: Trends and Applications
Автор: Lakshmi D, Ravi Shekhar Tiwari, Rajesh Kumar Dhanaraj
Издательство: CRC Press
Год: 2024
Страниц: 335
Язык: английский
Формат: pdf (true)
Размер: 20.9 MB

This book presents innovative research works to automate, innovate, design, and deploy Artificial Intelligence (AI) for Real-World Applications. It discussed AI applications in major cutting-edge technologies and details about deployment solutions for different applications for sustainable development. The application of Blockchain techniques illustrates the ways of optimisation algorithms in this book. The challenges associated with AI deployment are also discussed in detail, and edge computing with Machine Learning solutions is explained. This book provides multi-domain applications of AI to the readers to help find innovative methods towards the business, sustainability, and customer outreach paradigms in the AI domain.

• Focuses on virtual machine placement and migration techniques for cloud data centres
• Presents the role of Machine Learning and meta-heuristic approaches for optimisation in cloud computing services
• Includes application of placement techniques for quality of service, performance, and reliability improvement
• Explores data centre resource management, load balancing and orchestration using Machine Learning techniques
• Analyses Dynamic and scalable resource scheduling with a focus on resource management

Artificial Intelligence has rapidly permeated our lives, transforming industries, healthcare, and our interactions with technology. While AI promises innovation and efficiency, its opacity and complexity raise concerns about trust, accountability, and ethical implications. To address these challenges, Explainable AI (XAI) emerged, bridging the gap between machine learning and human understanding. XAI seeks to explain how AI systems make decisions, ensure their trustworthiness, and evaluate their societal and ethical impact.

This book provides a multifaceted realm of XAI, covering its theoretical foundations, practical applications, and ethical considerations. Through expert insights, real‑world examples, and accessible explanations, we demystify XAI’s complexities for a wide audience. Along the way, we explore the black box of AI models, learn various interpretation techniques, and examine responsible AI development and deployment. XAI’s influence extends from healthcare to finance, and autonomous vehicles to criminal justice, making it essential for various domains. Our aim is to empower readers with knowledge and insights to navigate this evolving field.

Skater is a free, open‑source model interpretation framework created for all models to develop a comprehensible ML model. It is a Python library which was designed to demystify the learned structure by the dataset in the black box model by globally referencing the dataset as well as by locally referencing the dataset. It implements LIME to validate the model decision policies for the single prediction, which uses the surrogate models to assess our model. It is a post hoc model interpretation algorithm.

ELI5 is an open‑source Python Unified Library, which is compatible with many Deep Learning frameworks that include Keras, CatBoost, LightGBM, XGBoost, Scikit‑learn, and sklearn‑crfsuite. It is based on Permutation Importance for explaining local and global interpretation of dataset and depends on LIME for interpreting and analyzing black‑box models.

Whether you are a student, researcher, practitioner, academic researchers in Computer Science and information technology, or simply curious, this book provides a solid foundation for comprehending Explainable AI.