The video discusses the intuition for semi-supervised labeling methods in Scikit-learn in Python.
Timeline
(no code)
00:00 - Outline of video
00:50 - Supervised, semi-supervised, unsupervised
04:06 - Assumptions
04:42 - Proportion of unlabeled data vs. classification accuracy
06:09 - Self training: One sense per collocation
11:22 - Label propagation
19:26 - Label spreading
19:42 - Code snippet
20:22 - Ending notes
