Название: Explainable AI Within the Digital Transformation and Cyber Physical Systems: XAI Methods and Applications
Автор: Moamar Sayed-Mouchaweh
Издательство: Springer
Год: 2021
Страниц: 201
Язык: английский
Формат: pdf (true), epub
Размер: 26.8 MB

This book presents Explainable Artificial Intelligence (XAI), which aims at producing explainable models that enable human users to understand and appropriately trust the obtained results. The authors discuss the challenges involved in making machine learning-based AI explainable. Firstly, that the explanations must be adapted to different stakeholders (end-users, policy makers, industries, utilities etc.) with different levels of technical knowledge (managers, engineers, technicians, etc.) in different application domains. Secondly, that it is important to develop an evaluation framework and standards in order to measure the effectiveness of the provided explanations at the human and the technical levels.

This book gathers research contributions aiming at the development and/or the use of XAI techniques in order to address the aforementioned challenges in different applications such as healthcare, finance, cybersecurity, and document summarization. It allows highlighting the benefits and requirements of using explainable models in different application domains in order to provide guidance to readers to select the most adapted models to their specified problem and conditions.

Machine learning methods, especially deep neural networks, are becoming increasingly popular in a large variety of applications. These methods learn from observations in order to build a model that is used to generalize prediction (classification or regression) to unknown data. Machine Learning methods generate or learn a highly effective mapping between input and output. Hence, they behave as “black box” entailing a huge difficulty for humans to understand how and why the prediction (output) was made. However in many applications, it is important to explain to users how the decision (prediction) was made by the model and its meaning using understandable terms. Indeed, explainable models allow users to trust them and to better use them thanks to the detailed information (explanation) on how and why they arrived to the provided prediction. Therefore, making machine learning models transparent to human practitioners or users leads to new types of data-driven insights.

XAI aims at producing explainable models that enable human users to understand and appropriately trust the obtained results. The produced explanations allow to reveal how the model functions, why it behaves that way in the past, present and future, why certain actions were taken or must be taken, how certain goals can be achieved, how the system reacts to certain inputs or actions, what are the causes for the occurrence of a certain fault and how this occurrence can be avoided in the future, etc.

- Includes recent developments of the use of Explainable Artificial Intelligence (XAI) in order to address the challenges of digital transition and cyber-physical systems;
- Provides a textual scientific description of the use of XAI in order to address the challenges of digital transition and cyber-physical systems;
- Presents examples and case studies in order to increase transparency and understanding of the methodological concepts.