An online learning technique for coping with novelty detection and concept drift in data streams
暂无分享,去创建一个
To learn time-dependent concepts from streams of examples is one of the greatest challenges in machine learning. The ability to identify novel concepts as well as to deal with concept drifts are two important tasks in this scenario. This paper presents a technique based on the k-means clustering algorithm that treats these tasks as parts of a single learning strategy. Early experimental results are discussed and they inspire further investigation.
[1] Vic Barnett,et al. Outliers in Statistical Data , 1980 .
[2] Michèle Basseville,et al. Detection of abrupt changes: theory and application , 1993 .
[3] Catherine Blake,et al. UCI Repository of machine learning databases , 1998 .
[4] S. Marsland. Novelty Detection in Learning Systems , 2008 .