En
Register Log in

List of publications on a keyword: «потоковые данные»

  • Adaptive real-time machine learning: methods for combating concept deviation in flow data

    Proceeding
    Education in the 21st Century: Challenges, Practices, and Development Trends
    Authors:
    Aleksandr S. Prudkii, Жевлакова Кристина Евгеньевна, Vladislav V. Kukartsev
    Work direction:
    Conditions of the pre-school educational environment that create and ensure the safety of pupils
    Abstract:
    The article explores modern approaches to adaptive machine learning in real-time environments, focusing on the problem of concept drift in streaming data. It examines the main methods for detecting and correcting data drift, which allow models to maintain accuracy when working with non-stationary information streams. The article analyzes online learning algorithms, adaptive threshold mechanisms, and incremental model update techniques. It also discusses the potential applications of these methods in monitoring, forecasting, and automation systems, as well as the challenges associated with computational complexity and processing delays.
    Keywords:
    adaptive machine learning, streaming data, concept rejection

  • Adaptive real-time machine learning: methods for combating concept deviation in flow data

    Proceeding
    Modern educational process: psychological and pedagogical support, educational strategies
    Authors:
    Aleksandr S. Prudkii, Жевлакова Кристина Евгеньевна, Vladislav V. Kukartsev
    Work direction:
    Varia (принимаются материалы по другим направлениям, соответствующим теме конференции)
    Abstract:
    The article explores modern approaches to adaptive machine learning in real-time environments, focusing on the problem of concept drift in streaming data. It examines the main methods for detecting and correcting data drift, which allow models to maintain accuracy when working with non-stationary information streams. The article analyzes online learning algorithms, adaptive threshold mechanisms, and incremental model update techniques. It also discusses the potential applications of these methods in monitoring, forecasting, and automation systems, as well as the challenges associated with computational complexity and processing delays.
    Keywords:
    adaptive machine learning, streaming data, concept rejection