Abstract | ||
---|---|---|
A data stream is a potentially uninterrupted flow of data. Mining this flow makes it necessary to cope with uncertainty, as only a part of the stream can be stored. In this paper, we evaluate a statistical technique which biases the estimation of the support of patterns, so as to maximize either the precision or the recall, as chosen by the user, and limit the degradation of the other criterion. Theoretical results show that the technique is not far from the optimum, from the statistical standpoint. Experiments performed tend to demonstrate its potential, as it remains robust even under significant distribution drifts. |
Year | DOI | Venue |
---|---|---|
2007 | 10.1016/j.patcog.2006.03.006 | Pattern Recognition |
Keywords | Field | DocType |
data streams,precision,accuracy.,significant distribution drift,recall,theoretical result,statistical technique,uninterrupted flow,frequent pattern,statistical standpoint,data stream,concentration inequalities,data mining,accuracy | Data mining,Data stream mining,Data processing,Data stream,Flow (psychology),Artificial intelligence,Recall,Machine learning,Mathematics,Data flow diagram | Journal |
Volume | Issue | ISSN |
40 | 2 | Pattern Recognition |
Citations | PageRank | References |
6 | 0.46 | 26 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Pierre-alain Laur | 1 | 21 | 4.66 |
Richard Nock | 2 | 1056 | 103.66 |
Jean-Emile Symphor | 3 | 23 | 5.08 |
Pascal Poncelet | 4 | 768 | 126.47 |