A New On-Line Learning Method for Coping with Recurring Concepts: The ADACC System

摘要

When the environment changes, as is increasingly the case when considering unending streams and long-life learning tasks, it is necessary to rely on on-line learning with the capability to adapt to changing conditions a.k.a. concept drifts. Previous works have focused on means to detect changes and to adapt to them. Ensemble methods relying on committees of base learners have been among the most successful approaches. In this paper, we introduce a new second-order learning mechanism that is able to detect relevant states of the environment in order to recognize recurring contexts and act pro-actively to concepts changes. Empirical comparisons with existing methods on well-known data sets show the advantage of the proposed algorithm.