Skip to main content
Fachgebiete
Automobil + Motoren Bauwesen + Immobilien Business IT + Informatik Elektrotechnik + Elektronik Energie + Nachhaltigkeit Finance + Banking Management + Führung Marketing + Vertrieb Maschinenbau + Werkstoffe Versicherung + Risiko
DE
EN
Themenseiten
Organisationspsychologie Projektmanagement Marketing Smart Manufacturing
Bücher
Zeitschriften
Weitere Formate
Podcasts Webinare Technik Webinare Wirtschaft Kongresse Veranstaltungskalender Awards
Jetzt Einzelzugang starten
Zugang für Unternehmen
Referenzkunden
MTZ-Fachkongress

Antriebe und Energiesysteme von morgen


Austausch von Automobil- und Energiewirtschaft

Am 14./15. Mai geht es in Chemnitz um die Mobilitätswende durch Elektro- und Wasserstofffahrzeuge.
Erweiterte Suche
Anmelden
nach oben

2024 | Buch

Ambient Intelligence (AmI) Kapitel lesen Erstes Kapitel lesen

Explainable Ambient Intelligence (XAmI)

Explainable Artificial Intelligence Applications in Smart Life

verfasst von: Tin-Chih Toly Chen

Verlag: Springer Nature Switzerland

Buchreihe : SpringerBriefs in Applied Sciences and Technology

Enthalten in: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Einloggen, um Zugang zu erhalten
insite
SUCHEN

Über dieses Buch

This book systematically reviews the progress of Explainable Ambient Intelligence (XAmI) and introduces its methods, tools, and applications.

Ambient intelligence (AmI) is a vision in which an environment supports the people inhabiting it in an unobtrusive, interconnected, adaptable, dynamic, embedded, and intelligent way. So far, artificial intelligence (AI) technologies have been widely applied in AmI. However, some advanced AI methods are not easy to understand or communicate, especially for users with insufficient background knowledge of AI, which undoubtedly limits the practicability of these methods. To address this issue, explainable AI (XAI) has been considered a viable strategy. Although XAI technologies and tools applied in other fields can also be applied to explain AI technology applications in AmI, users should be the main body in the application of AmI, which is slightly different from the application of AI technologies in other fields.

This book containsreal case studies of the application of XAml and is a valuable resource for students and researchers.

Inhaltsverzeichnis

Frontmatter
Chapter 1. Ambient Intelligence (AmI)
Abstract
This chapter begins by defining ambient intelligence (AmI). The chronology of the emergence of the most widely used AmI technologies is then provided. Through reviewing some cases in the literature, AmI applications prevalent in various fields are mentioned, such as emotionally pleasing design, telemedicine and telecare, context-aware recommendation, home care and assisted living, smart home, smart tourism, and smart factory. The issues and challenges that need to be addressed by existing AmI applications are then discussed. To solve these problems and challenges, the most popular artificial intelligence (AI) technologies are highlighted. However, some AI technologies are very complex, limiting the interpretability and credibility of related AmI applications. To address this problem, the combination of AmI and XAI results in XAmI, and a review of some existing XAmI applications is performed. Based on the review findings, the application of XAmI techniques and tools is expected to mitigate the social impacts of AI applications in AmI on fairness and bias, security, verifiability, and accountability.
Tin-Chih Toly Chen
Chapter 2. Explainable Artificial Intelligence (XAI) with Applications
Abstract
This chapter begins by defining explainable artificial intelligence (XAI). The explainabilities of existing ML models are then compared, showing the need for XAI applications to improve the explainabilities of some ML models. To this end, XAI techniques and tools for interpreting and enhancing artificial intelligence applications, especially in the field of ambient intelligence (AmI), are discussed. Subsequently, the requirements for trustable AI and XAI are listed. Various classifications of existing XAI methods are then performed to meet these requirements. A literature analysis was also conducted on the application of XAI in various fields, showing that services, medicine, and education are the most common application fields of XAI. Finally, several types of explanations are introduced. XAI techniques for feature importance evaluation are also described.
Tin-Chih Toly Chen
Chapter 3. XAmI Applications to Smart Homes
Abstract
A smart home is an environment where users spend most of their time. However, unlike location-aware systems or telemedicine or telecare systems, smart homes can be controlled by the user, even from the outside. Therefore, assisting users in operating smart home appliances has become a key task. To accomplish this task, various sensors and actuators are installed in smart home appliances to detect user conditions and needs, accompanied by many artificial intelligence (AI) applications. This chapter first summarizes the applications of AI technologies in smart homes. Explainable ambient intelligence (XAmI) techniques can then be applied to extract useful behavioral patterns of users from these AI applications, thereby making such AI applications more autonomous and intelligent. Common XAmI techniques for this purpose include fuzzy inference rules, decision rules (or trees), random forests, and locally interpretable model-agnostic explanations (LIME).
Tin-Chih Toly Chen
Chapter 4. XAmI Applications to Location-Aware Services
Abstract
Location-aware (or location-based) services (LASs) are probably the most prevalent application of ambient intelligence (AmI). This chapter first summarizes the applications of artificial intelligence (AI) in LASs. However, some AI applications in LASs are difficult to understand or communicate with mobile users, so explainable ambient intelligence (XAmI) techniques must be applied to enhance the understandability of such AI applications. To this end, some XAmI techniques for LASs have been introduced. Subsequently, various types of interpretations of LASs are explained through examples. In addition, the priorities and impacts of criteria for choosing suitable service locations are also distinguished. Furthermore, to enhance the interpretability of the recommendation process and result for users, visualization XAmI methods tailored for LASs are also reviewed. This section concludes with a discussion of how to interpret AI-based optimization in LASs.
Tin-Chih Toly Chen
Chapter 5. XAmI Applications to Telemedicine and Telecare
Abstract
Telemedicine and telecare are another important application of ambient intelligence (AmI). This chapter first summarizes the applications of artificial intelligence (AI) in telemedicine and telecare. Since some of these AI applications are difficult to understand or communicate with patients, various explainable ambient intelligence (XAmI) techniques have been applied, such as shape-added explanation value (SHAP) analysis and locally interpretable model-agnostic explanation (LIME) to overcome such difficulties. Telemedicine services for type-II diabetes diagnosis are taken as an example to illustrate such applications. Several issues with existing XAmI applications in telemedicine and telecare are then discussed. It is worth noting that after SHAP analysis, some important attributes may be difficult to measure by patients themselves, which affects the utility of telemedicine or telecare applications.
Tin-Chih Toly Chen
Metadaten
Titel
Explainable Ambient Intelligence (XAmI)
verfasst von
Tin-Chih Toly Chen
Copyright-Jahr
2024
Electronic ISBN
978-3-031-54935-9
Print ISBN
978-3-031-54934-2
DOI
https://doi.org/10.1007/978-3-031-54935-9

Premium Partner

    Bildnachweise