Machine Learning (ML) and other artificial Intelligence (AI) techniques have been developed for real-time decision making, and are gaining traction in data-rich situations. However, these techniques are less proven in sparse-data environments, and at present are more the subject of research than application. Typical implementations of ML and AI require a cross-disciplinary decision engine that, once “trained,” can cognitively respond to changes in input. The key to successful training is to a) have a defined decision-basis (answer-key), and/or b) facilitate sufficient learning, both of which require ample data (observability) and ample time for the machine to develop a logical outcome. Much research has been focused on developing decision algorithms using various logical formulations, dimensionality reductions, neural techniques, and learning reinforcements for tasks that traditionally require human intelligence. What is missing in most current research streams are implementations of ML and AI for decisions that are fundamentally rooted in human intuition and empathy, e.g., situations in which the decision requires a holistic view and the outcome is based on a qualitative judgement based on context and fact. This paper is intended to benefit a wide range of readers considering Artificial Intelligence, from the merely curious to “techies” from other disciplines to experienced practitioners and researchers. Using a qualitative/ characteristics base perspective of data and AI, we examine defense industry procurement, operational, tactical, and strategic decision scenarios, then identify where AI can currently promote better informed decisions and which arenas need would benefit by letting AI technology and sophistication evolve further.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Artificial Intelligence and Machine Learning in Sparse/Inaccurate Data Situations


    Beteiligte:
    Dixit, Rahul (Autor:in) / Chinnam, Ratne Babu (Autor:in) / Singh, Harpreet (Autor:in)


    Erscheinungsdatum :

    01.03.2020


    Format / Umfang :

    4348003 byte




    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch




    On Learning From Inaccurate and Incomplete Traffic Flow Data

    Alesiani, Francesco / Moreira-Matias, Luis / Faizrahnemoon, Mahsa | IEEE | 2018



    Artificial intelligence and machine learning

    Kühl, Niklas / Schemmer, Max / Goutier, Marc et al. | Online Contents | 2022

    Freier Zugriff

    Machine Learning and Artificial Intelligence

    Sadatuzzaman Saagoto, Md. / Farhan, Ahmed / Barua, Protik et al. | Springer Verlag | 2024