I then compared those extractions to one another using a statistical method of dimensionality reduction called t-distributed stochastic neighbor embeddings.

然后，我使用一种称为 t 机邻域入的降方法将这些提取结果相互进行比较。