在一些实际问题中,数据的分布随时间的变化而逐渐变化,这类数据的学习问题被称之为演化数据的学习.文中综述了演化数据上的学习方面的研究进展.提出了今后需要关注的一些问题,如数据演化的机制、一般性的假设问题、演化数据分类等等.
In some real applications, the data distributions often evolve over time. Machine learning on this kind of data is named learning on time-evolving data. This paper surveys the re cent advance of the research on this topic, and introduces some problems worth attention in the future, such as the mechanism of Time-Evolving Data, generality, classification.