如果不涉及"learning",算法是否可以归类为"unsupervised learning"?

Can an algorithm be classified as "unsupervised learning" if there is no "learning" involved?

基本上,我的问题是,由于无监督学习是一种机器学习,是否需要机器的某些方面 "learning" 并根据其发现进行改进?例如,如果开发了一种算法来获取未标记的图像并找到它们之间的关联,它是否需要根据这些关联改进自身以被分类为 "unsupervised learning",或者仅仅报告这些关联足以获得该分类?

For example, if an algorithm is developed that takes unlabeled images and finds associations between them...

那是 "unsupervised learning," 中的 "learning" 所以是的,这将被视为无监督学习。

...does it need to improve itself based on those associations...

不,不要求算法利用它所学的知识并自我改进才能被视为无监督学习。只需分析数据集并找到以前未知的关联就足以被视为无监督机器学习。 "unsupervised" 区别实际上只是初始数据集未标记。