| 摘要 |
<p>Known methods for providing data for modelling a time-variant behaviour are either inherently inaccurate or become inaccurate when a change takes place in the time-variant behaviour. A method of the rationalisation of data used to model a time-variant behaviour is disclosed herein. This provides the advantages that the storage requirements for such data are reduced and that the accuracy of detection of events in the behaviour is increased. The method sees the addition of labels to training data that indicates whether that data relates to recent events or not. A classifier is generated from the labelled training data. By removing old data which the classifier would classify differently were the old data re-labelled as new, a selective purging of the old training data takes place each time new training data becomes available. The method is especially useful in detecting fraudulent use of, or faults in, a communications network. <IMAGE></p> |
