Drift learning for decision-making in dynamic multi-stream environments

动态多流环境中决策的漂移学习

• 批准号: DP190101733
• 负责人: A/Prof Guangquan Zhang
• 金额: $ 32.8万
• 依托单位: University of Technology Sydney
• 依托单位国家: 澳大利亚
• 项目类别: Discovery Projects
• 财政年份: 2019
• 资助国家: 澳大利亚
• 起止时间: 2019-06-19 至 2022-09-30
• 项目状态: 已结题
• 来源: https://dataportal.arc.gov.au/RGS/Web/Grants/DP190101733
• 关键词: Drift; learning; decision; making; dynamic

This project aims to provide application-ready real-time decision support systems for big data situations. Real-time support for organisational decisions is crucial in fast-changing environments that are highly dependent on data from multiple large streams. Unforeseen changes in data distribution (drift) are inevitable. The ability to learn drift in dynamic environments with multiple large data streams will benefit innovation and decision quality in challenging data situations. The project will have wide applications, such as in cybersecurity, telecommunications, bushfire control and logistics. The project will advance machine learning knowledge, providing a foundation and technologies to support real-time decision-making in big data environments.

该项目旨在为大数据情况提供应用程序就绪的实时决策支持系统。在高度依赖来自多个大型数据流的快速变化的环境中，对组织决策的实时支持至关重要。数据分布中不可预见的变化（漂移）是不可避免的。在具有多个大型数据流的动态环境中学习漂移的能力将有利于在具有挑战性的数据情况下的创新和决策质量。该项目将有广泛的应用，如网络安全，电信，消防控制和物流。该项目将推进机器学习知识，为支持大数据环境中的实时决策提供基础和技术。