Concept Drift Detection and Reaction for Data-driven Decision Making

所有组织数据和非结构化数据都会发生不可预见的数据模式变化(概念漂移)，随着时间的推移，使得后续数据驱动的预测变得不那么准确，从而导致糟糕的决策结果。为了解决这些问题，本项目旨在开发新的模糊能力模型来反映概念漂移，以及检测和响应变化的方法，并将它们集成到决策支持系统(DSS)中，以提供对不断变化的环境的适应性。这些前沿成果旨在直接用于加强组织的实时数据分析和动态决策，并有望通过引入一个新的研究领域--适应性数据驱动的决策支持系统，对信息科学作出重大贡献。

Unforeseeable changes to patterns that underlie data (concept drift) occur in all organisational data, and in unstructured data, making subsequent data-driven prediction less accurate as time passes, which leads to poor decision outcomes. To solve these problems, this project aims to develop novel fuzzy competence models to reflect concept drift, with methods to detect and react to changes, and integrate them into Decision Support Systems (DSS) to provide adaptivity for ever-changing environments. These cutting-edge results are intended to be directly used to enhance organisational real-time data analytics and dynamic decision making, and are expected to significantly contribute to information science by introducing a new research field, adaptive data-driven DSS.