摘要 |
A sample data reliability evaluation method includes a first calculating step of calculating a representative distance between samples of a same cluster, which is a representative distance between the sample and the sample belonging to the same cluster to that to which the sample belongs, a second calculating step of calculating a representative distance between samples of different clusters, which is a representative distance between the sample and the sample belonging to the different cluster from that to which the sample belongs, a third calculating step of calculating a reliability evaluation index, which is an index for evaluating reliability of the sample data based on the representative distance between samples of the same cluster and the representative distance between samples of different clusters, and an evaluating step of evaluating the reliability of a subset of the sample data set or the whole sample data set.
|