Название: Explainable Artificial Intelligence: A Practical Guide
Автор: Parikshit Narendra Mahalle, Yashwant Sudhakar Ingle
Издательство: River Publishers
Год: 2024
Страниц: 104
Язык: английский
Формат: pdf (true), epub
Размер: 14.6 MB

This book explores the growing focus on Artificial Intelligence (AI) systems in both industry and academia. It evaluates and justifies AI applications while enhancing trust in AI outcomes and aiding comprehension of AI feature development. Key topics include an overview of explainable AI, black box model understanding, interpretability techniques, practical XAI applications, and future trends and challenges in XAI.

Technical topics discussed in the book include:
* Explainable AI overview
* Understanding black box models
* Techniques for model interpretability
* Practical applications of XAI
* Future trends and challenges in XAI

Explainable Artificial Intelligence (XAI) comprises a set of frameworks and tools to assist us in forecasting futuristic events with the aid of Machine Learning/evolutionary and intelligent techniques. XAI helps to improve the performance of the automated models and to train the automated tools for diverse engineering purposes. XAI can also assist in the generation of feature attributions for forecasting the model behavior with respect to different inputs. XAI is used in diverse fields such as marketing, data science, engineering, medical science, and economics. All these fields use XAI-enabled tools to identify gaps in data, determine the biases in the trained models, and check whether the trained models are drifting towards a particular type of data. The outcomes of XAI need transparency to align the output with human-interpretable explanations. The proposed book attempts to cover research work and use cases based on XAI for building interpretable tools to grow end-user trust and to improve the performance of models based on explainable AI.

Explainable Artificial Intelligence: A Practical Guide is a comprehensive guide to the reader which covers the fundamentals of traditional AI to the current status of XAI. The first chapter provides a basic overview of XAI along with its importance. The second chapter provides key limitations of Machine Learning as a black box and puts forth the need for XAI. The third chapter of this book focuses on model interpretability through popular techniques available in state of the art. The fourth chapter describes the key applications of explainable AI in various fields which includes healthcare, finance, autonomous vehicles, recommender systems and agriculture. The last chapter of this book covers the outlook of explainable AI along with the key challenges for design and development and concludes the book.

In a nutshell, this book puts forward the best research roadmaps, strategies and challenges to design and develop XAI applications. The book will motivate readers to use this technology for better analysis of the needs of layman and educated users to design various use cases in XAI. The book is useful for undergraduates, postgraduates, industry, researchers and research scholars in ICT, AI, Machine Learning. We are sure that this book will be well-received by all stakeholders.