1. When we assess objects of interest in these domains (e.g., in terms of their correctness and quality), we often resort to subjective evaluation and provide rating information that is typically imprecise. Therefore, ordinal labels are preferred to continuous scores.
by mean of 用, 依靠
1. After finding all these neighbors, we need to aggregate their class labels by means of a function
1. When we assess objects of interest in these domains (e.g., in terms of their correctness and quality), we often resort to subjectiveevaluation and provide rating information that is typically imprecise. Therefore, ordinal labels are preferred to contin