| 摘要 |
<p>The invention relates to a method of evaluating a measured behaviour against a predicted behaviour, where the measured behaviour is supported by a set of first data and the predicted behaviour is predicted from a set of second data. The evaluation is primarily used to provide an explanation for differences between measured and predicted behaviours. The method comprises the steps of: i) comparing the measured behaviour with at least one aspect of the predicted behaviour; ii) if there is a difference between the measured behaviour and said at least one aspect, which difference meets one or more predetermined criteria, using the set of first data together with at least some of the set of second data to generate a further behaviour; iii) comparing the further behaviour with the measured behaviour; and iv) if there is a difference between the further behaviour and the measured behaviour, which difference satisfies one or more predetermined criteria, reviewing the set of first data against the set of second data to identify one or more elements of the first data that are relevant to said difference. <IMAGE></p> |
