| 摘要 |
The present invention provides a system and method for automatically deriving unique surrogate response data from experiment results in which inherent data loss occurs in a sufficient number of the samples to disallow quantitative effects estimation at the experimenter's desired level of confidence for statistical significance. In part, the unique surrogate response data sets of the present invention have four primary characteristics including: each is numerically analyzable; each may be more readily or directly obtained in which inherent data loss occurs; each provides a response value for an experiment trial; and each provides information on the effect of the change made to the process or system that would have been obtainable if the experiment samples had had no inherent data loss.
|
