Anytime Learning - The next Step in Organic Computing?

Abstract

Anytime learning describes a relatively novel concept by which systems are able to acquire knowledge about a changing environment and adapt and behave accordingly to this. "Anytime" refers to the fact that the system is capable of returning imperfect results at any point in time, which allows it to remain functional even if a perfect solution could not be found within the necessary time frame. This paper focuses on illustrating the concept of anytime learning and examining how it relates to organic computing as a whole. Could anytime learning be the next step in organic computing?