半监督分类问题的一个直观示例说明 (An example of the use of semi-supervised classification)

The semi-supervised learning works by making the so-called cluster assumption: the classification rule does not change in regions of the input space that are densely populated. In other words, the algorithm chooses a classification decision boundary that lies in a region of low density (as Figure 1). More details can be found in paper (Protein Ranking by Semi-Supervised Network Propagation).