An online classification system maintained by a learner can besubject to latency and filtering of training examples which can impact on itsclassification accuracy especially under concept drift. A life-cycle model isdeveloped to provide a framework for studying this problem. Meta dataemerges from this model which it is proposed can enhance online learningsystems. In particular, the definition of the time-stamp of an example, ascurrently used in the literature, is shown to be problematic and an alternative isproposed.
|Title of host publication||Unknown Host Publication|
|Number of pages||8|
|Publication status||Published - 24 Sep 2010|
|Event||HaCDAIS 2010: International Workshop on Handling Concept Drift in Adaptive Information Systems: Importance, Challenges and Solutions - European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD 2010) Barcelona|
Duration: 24 Sep 2010 → …
|Workshop||HaCDAIS 2010: International Workshop on Handling Concept Drift in Adaptive Information Systems: Importance, Challenges and Solutions|
|Period||24/09/10 → …|